Policy talk:Terms of Use/Archive 2

From Wikimedia Foundation Governance Wiki

20 February 2023: Start of the Terms of Use update process

The m:Wikimedia Foundation Legal department is hosting a feedback cycle about updating the Wikimedia Terms of Use (ToU) from February 21 to April 24, 2023. Feedback regarding the proposed update are welcome in any language below this message.

On behalf of the Legal Department,
RamzyM (WMF) (talk) 10:53, 20 February 2023 (UTC)

Minor copyedits

While the ToU are already being updated, there are a two minor changes I believe should made:

  • Section 7c says you must credit the author(s) in a reasonable fashion if you import text under a CC BY-SA license that requires attribution. All CC BY-SA licenses require attribution, but some licenses compatible with CC BY-SA do not require attribution. Is this what this clause is referring to? In any case, I think its meaning should be clarified. There is also the question of whether we wish to require attribution regardless of the legal requirement to do so, but that is a more substantive change.
  • Section 16 says in at least three languages. Should this be in at least three (3) languages?

Thank you for all the hard work you are putting into this process! Best, HouseBlaster (talk) 20:03, 20 February 2023 (UTC)

The line that would result in:
  • Engaging in ,threats, stalking, spamming, vandalism or harassment as described in the UCoC;
Should probably be:
  • Engaging in threats, stalking, spamming, vandalism, or harassment as described in the UCoC;
— xaosflux Talk 15:44, 21 February 2023 (UTC)
Can confirm that the proposed change is in line with the original copy. Fixed. RamzyM (WMF) (talk) 22:40, 21 February 2023 (UTC)
1 Srbrazuk01 (talk) 14:48, 24 February 2023 (UTC)
  • Section 4: to binding “Med-Arb” (a “Marketing Company Mediation”). As described should be to binding “Med-Arb” (a “Marketing Company Mediation”), as described. HouseBlaster (talk) 17:58, 21 February 2023 (UTC)

"Part of our mission is to we offer websites and technical infrastructure to help you do this." – That's not a sentence. So why is the third bullet point part of the list? Simply remove the bullet point from "We offer websites..." to fix the issue. ToBeFree (talk) 18:28, 22 February 2023 (UTC)

Another minor thing: obvious it is the intention to change "users like yourself" to just "users". But the wording "like yourself" is still found elsewhere in the text. Ziko (talk) 22:10, 22 February 2023 (UTC)
Another, very very minor thing: „Content that is in the public domain is welcome!“ Using an exclamation mark is rather unusual in such a text. Ziko (talk) 22:36, 22 February 2023 (UTC)
  • Section 7(c): compatible with CC BY-SA (or, as explained... should be compatible with CC BY-SA 4.0 (or, as explained..., as the abbreviation for "Creative Commons Attribution-ShareAlike 4.0 International" was defined to be "CC BY-SA 4.0" (I agree that defining the version number in the abbreviation is an improvement). HouseBlaster (talk) 21:30, 25 February 2023 (UTC)

Section 10 contains a sentence fragment: Please note that these mailboxes are monitored by users of the Projects, not the Foundation. So, so they should not be threatened or issued with legal demands. "Issued with" also reads weird to me, but I think that's an English dialect thing. 3mi1y (talk) 07:39, 2 March 2023 (UTC)

I think a coma is required: As used throughout the rest of the terms, our services consist of. For clarity, we could also add quotes arround “our services”. --Pols12 (talk) 15:54, 15 March 2023 (UTC)

More commas suggested: (1) one under "No Harm", following "You do not harm our technology infrastructure", lest someone interpret the new second clause to be "You do not ... follow the policies for that infrastructure"; and (2) another under "Our Terms of Use", between the new "San Francisco, California" and the following "whose mission...", as "California" is parenthetical there and should be set off by paired commas. Also: why do sections 14, 15 and 16 have a pink background? If that's supposed to indicate something important, what is it? If not, remove it. — JohnFromPinckney (talk / edits) 07:12, 21 April 2023 (UTC)

Insertion of "encyclopedic" in Section 1 a)

Changes in Section 1 a):

Because the Wikimedia Projects are collaboratively edited, the vast majority all of the encyclopedic content that we host is provided by users like yourself, and we do not take an editorial role.

Unless you intend to take an editorial role on Commons, Wikidata, Wikivoyage, Wikisource, Wikiqoute ..., I'd like to suggest reconsidering the addition of "encyclopedic" here. El Grafo (talk) 13:36, 21 February 2023 (UTC)

Ack, I was confused myself by this addition. Best, —DerHexer (Talk) 16:04, 22 February 2023 (UTC)
Yes, me too. What is meant here is the "main content", different from the "auxiliary content" or "meta content" (talk pages, rule pages, election pages etc.). I think there is no established term; for my book I had to make one up. (It is not the same as the "main namespace", but close.) Ziko (talk) 22:13, 22 February 2023 (UTC)
Yes - I think "content that it is the goal of the project to create" is what is meant, but there has to be a more elegant way of saying that (FWIW special:statistics defaults to "content pages"). However, I'm not sure its necessary as the vast majority of all the content is created by users. Thryduulf (talk: meta · en.wp · wikidata) 01:01, 2 March 2023 (UTC)
Ideed. It makes sense to move away from the absolute "all content", as we do import content from elsewhere (especially on Commons). But the "WMF does not take an editorial role" applies equally outside the main content namespaces. I'd suggest to just use "the vast majority of all content". El Grafo (talk) 07:53, 2 March 2023 (UTC)
Agree: skip the word "encyclopedic". There are wikis other than wikipedia. Maybe change to
Because the Wikimedia Projects are collaboratively edited, most of the content that we host is provided by users like yourself. We do not take an editorial role.
Taylor 49 (talk) 08:28, 5 March 2023 (UTC)
Just confirming that the word encyclopedic will be committed from the next draft SSpalding (WMF) (talk) 17:35, 6 March 2023 (UTC)
Huh? "omitted", "included"??? Johnbod (talk) 17:42, 20 April 2023 (UTC)

Third Party Resources Policy

Section 4 adds a new reference to the Third Party Resources Policy. Is this a new policy or something existing that I've never noticed? In general the meaning of the term uploading third-party technical resources is not clear to me. Taavi (talk!) 14:47, 21 February 2023 (UTC)

Came here for this, only thing I came up with on search was an incomplete phab task, phab:T296847. — xaosflux Talk 15:17, 21 February 2023 (UTC)
@Samuel (WMF): can perhaps provide more on this? — xaosflux Talk 15:18, 21 February 2023 (UTC)
Hello @Taavi and Bawolff:. The Third Party Resources policy is a work in progress led by the Foundation’s Security team. It is aimed at formalizing the use of third-party resources by gadgets and user scripts, so as to ensure better privacy and security for users. Here, “third party resources” should be viewed as ​​computer resources which are located outside Wikimedia production websites. They may include but are not limited to: executable scripts, style sheets, image and font files, JSON/JSONP data.
While this is still ongoing work, the Third Party Resources policy is expected to complement the ToU by covering specific aspects such as the risks related to user scripts and gadgets loading third-party resources, best practices for script developers and gadget makers, administrative and technical enforcement.
So far, the work on the Third Party Resources policy has mainly been discussed in Phabricator (phab:T296847). As per the latest updates, a draft of the policy was crafted and the Security team has been reaching out to a number of community members, including Interface administrators, to gather initial community insights and adjust the draft before engaging in a much larger public discussion on the policy. Once this phase is complete, the policy draft will be shared publicly for discussion.
For the time being, please feel free to subscribe to the Phabricator ticket (phab:T296847) for updates on this ongoing work. -- Samuel (WMF) (talk) 11:32, 22 February 2023 (UTC)
I'm not sure I like the wording of this section if that is the case. Its not really the running or uploading of third party software that this is trying to solve (At least i think), but its about communicating data (potentially including PII - certainly if we call IP addressess PII) to third parties and protecting the integrity of the browsing experience from third parties. If anything, the desired outcome is not to ban uploading the third party resources but to encourage it - Since running the software locally plugs the privacy risk [Sort of, somewhat, I'm over-simplifying here]. Bawolff (talk) 11:42, 22 February 2023 (UTC)
+1, upload does not seem like the correct term to use when we're talking about adding references to external sites. Taavi (talk!) 15:57, 22 February 2023 (UTC)
Also, does this really need to be in ToU. It seems oddly specific. For comparison, we wouldn't include the gerrit +2 policy in the terms of use. Bawolff (talk) 19:30, 24 February 2023 (UTC)
According to [1], there are no plans to publish the Third party resources policy, even in draft form, prior to the end of this consultation. In my opinion, having a section of the proposed terms of use essentially be To be determined at a later date, is not an acceptable form of consultation. I would like to respectfully suggest that the "Unauthorized Uploading of Third-Party Technical Resources" section of the proposed changes be dropped, until such time such a policy (or at least a public draft of one) exists. Bawolff (talk) 00:16, 8 March 2023 (UTC)
I agree that this should be removed under the same conditions. Izno (talk) 04:36, 8 March 2023 (UTC)

Further feedback

Der-Wir-Ing ("DWI") talk 15:47, 21 February 2023 (UTC)

@Der-Wir-Ing - regarding missing links, these have often been left out to make wiki-syntax with regards to translations less complex. They will be included in final versions. DBarthel (WMF) (talk) 18:33, 21 February 2023 (UTC)
Yes, we will update the text with several links when we have it. Cheers, RamzyM (WMF) (talk) 00:20, 22 February 2023 (UTC)

Marketing Company Mediations in section 4

A minor copy editing note: the last sentence in this section is an incomplete sentence. I suggest it be made an explanatory clause of the previous sentence, by changing ". As described in section 14 of these Terms of Use." to ", as described in section 14 of these Terms of Use." isaacl (talk) 17:07, 21 February 2023 (UTC)

This one is copied verbatim from the original copy that we received for upload to Meta, but your suggestion is well noted. Thanks! RamzyM (WMF) (talk) 22:44, 21 February 2023 (UTC)

Section 4 says:

In addition, if you make a public posting on a third-party service advertising editing services on Wikipedia in exchange for pay, you must disclose what Wikipedia accounts you will use for this service in the public posting on the third-party service.

Should this be less specific? Something like:

In addition, if you make a public posting on a third-party service advertising editing services on Wikipedia in exchange for pay compensation of any kind, you must disclose what all Wikipedia account(s) you will use for this service in the public posting on the third-party service.

It might also be a good idea to prohibit logged-out paid editing, as it makes disclosure a nightmare. Thanks! HouseBlaster (talk) 18:06, 21 February 2023 (UTC)

This is all great, but something similar has existed for some time on en. In that time I have noticed that almost no editors have done this, and that there seems to be no enforcement of this by the WMF on third-party sites - which have no need to follow our policies anyway. Given that this has been absolutely useless to date, why add it to the ToU? If you have no power or wish to enforce it, why include it? - Bilby (talk) 13:44, 1 March 2023 (UTC)

There is another section in this discussion called "Marketing Company Mediations" where the question of why this update is important is addressed. The short answer though is that the "rules" are not being updated. Instead, we are making changes to get a much faster and cheaper way of actually enforcing the existing rules off-wiki. In short, we do currently have a strong interest to enforce the rules but for years we have realized that we don't have the capacity to do it realistically via litigation. The proposed "mediation" process is much more realistic that litigation to force bad actors to change. Said differently, our proposed update is a response to your exact concern. SSpalding (WMF) (talk) 15:56, 1 March 2023 (UTC)

if you make a public posting on a third-party service advertising editing services on Wikipedia in exchange for pay, you must disclose what Wikipedia accounts you will use for this service in the public posting on the third-party service.

Is this including LinkedIn? And company websites? Or only freelancer, fiverr and similar? Thanks Cartago90 (talk) 18:17, 21 February 2023 (UTC)

  • IMHO if somebody is advertising Wikipedia services on LinkedIn or a company website it still needs to be disclosed. I'd go so far as saying it needs to be disclosed on a page that's readable to the public (or is public part of "disclosed"). Smallbones (talk) 23:12, 21 February 2023 (UTC)
    That is my reading, too. It says you must disclose "in the public posting on the third-party service" (emphasis mine). HouseBlaster (talk) 23:27, 21 February 2023 (UTC)

ملاحظات أولية

مرحبًا، قرأت هذا البند "هذا يعني أننا عادة لا تراقب أو نحرر محتوى مواقع المشاريع، ونحن لا نتحمل أي مسؤولية عن هذا المحتوى." كلمة عادة تعني أنه توجد حالات يُراقب فيها المحتوى وأجده مناقضًا لهذا البند "وهو يناقض هذه "انت مسؤول قانونا عن تعديلاتك ومساهماتك في مشاريع ويكيميديا " فكيف أكون حرا ومسؤولا ومراقبا في آن واحد؟ تحياتي Nehaoua (talk) 20:01, 21 February 2023 (UTC)

[for @Jrogers (WMF) and SSpalding (WMF):] "Hello, I have noticed these two statements: "Generally we do not contribute, monitor, or delete content...." and "....and have no responsibility for the operations of the website or its content.", the word “generally” means there are cases where the content is monitored, and I see that this contradicts with the statement " You are responsible for your own actions: You are legally responsible for your edits and contributions on Wikimedia Projects … ". How can I be free, responsible and monitored at the same time? Regards". RamzyM (WMF) (talk) 13:15, 1 March 2023 (UTC)

Please let me know if there is a better place to ask this question. I'm sure we all know that almost nobody reads ToUs or Terms of Service, e.g. a majority of people reading this have been on Google today, but when was the last time you read their Terms of Service? I say it's been at least 15 years for me. With paid editing at issue here, I don't think just having the proper legal words in the proper format is good enough though. Every company in the world who might want to advertise here should have some idea that paid editing is prohibited here, even before they read the ToU. That's a lot of people who have to read it before it's effective in keeping out undeclared paid editors. In a legal sense, I'd guess, that it would be more effective as well, if when you get in front of the mediator and the paid editing company says "Sorry, I just never knew ..." that the mediator thinks "Oh really?" So I think to be really effective the WMF needs to thoroughly publish this in many places, in many cases. I'd feel very much more enthusiastic in supporting this if somebody near the top gave us an idea of how they are going to publicize this. Not just the wording of the ToU, but the ideas behind it. The WMF can do this better than anybody else in the movement. Smallbones (talk) 23:38, 21 February 2023 (UTC)

+1 Skimel (talk) 22:38, 22 February 2023 (UTC)
I also wonder how to enforce edits that might be paid while undisclosed. Meta:Requests for CheckUser information requires probable cause, so if we always assume good faith, then it is unenforceable.--Jusjih (talk) 02:11, 28 February 2023 (UTC)
Some of what's here has been addressed lower on the page (Marketing Company Mediations) in greater detail. To @Smallbones point, rules are most useful when they can be enforced. With these updates, the rules are mostly the same but we are hoping to drastically increase our ability to enforce those rules properly (making UPE companies feel like it's more likely that we will actually enforce the rules). One of the first efforts we would take, should the new language be supported by the community, would be that the Legal department would send notices to companies that the community has already identified as being systematic rule violators to directly let them know about the changes and that they would be a target for future enforcement. SSpalding (WMF) (talk) 18:49, 28 February 2023 (UTC)
This approach was accepted on en.wp in November 2020 [2]> I have noticed that of the 285 accounts which I know of on Upwork - one of the major freelance sites advertising WP editing - a total of 1 person (an en.wp admin) has disclosed their WP account. When this was proposed on en.wp it was argued that WMF Legal would use this to stop paid editing, but as far as I can tell this has had no impact. Can you give a rough number of how many paid editors have been contacted since this was added to the en.wp policy over two years ago, and how many subsequently added this disclosure to their advertisements? - Bilby (talk) 14:05, 1 March 2023 (UTC)
Although for both practical and confidentiality reasons, I can't provide exact numbers, I can say that enforcement letters go out every week from the Foundation to providers of undisclosed paid editing services (since I started at the Foundation in 2020). These bad actors sometimes comply with our requests and sometimes do not. The problem is that if these bad actors do not comply voluntarily, then WMF's only recourse (currently) is to file a lawsuit. Lawsuits are expensive, time consuming, and resource intensive. That means that, currently, we may begrudgingly decide to not litigate even against the worst actors.
This change would give us a much more effective way of encouraging bad actors in these demand letters to stop. And if they don't, it gives us a cheap and effective way to force them into a formal legal process to compel them to. Said differently, this update is a response to increase the effectiveness of our current process because there are limitations with it currently. SSpalding (WMF) (talk) 16:08, 1 March 2023 (UTC)
I don't understand why you can't even state a number due to confidentiality. Confidentiality would, naturally, entail not stating who those letters were sent to, but stating how often you tried to enforce existing policies seems like a completely different issue. However, I am unsurprised that you will not provide numbers.
That said, can you provide a case where this would work? Let's put forward a simple case - Y has been advertising their services on Upwork, and have picked up over 200 paid editing jobs on WP. All are undisclosed, and each time they have created a new account or used an IP address to make changes. You do not know their full name, address, or email. How is this to be enforced? - Bilby (talk) 20:05, 1 March 2023 (UTC)
Thinking about it, that might be unfair as it would be unenforceable anyway. How about I modify it. You are able to identify the Upwork contractor, and so have their name and are able to find contact details. You've contacted them and they have failed to add their WP details to their Upwork account. Based on what you said, it was not viable to litigate in order to force them to provide those details or to stop editing WP. So how do the proposed changes help? - Bilby (talk) 22:28, 1 March 2023 (UTC)

Chilling Effects website

https://www.chillingeffects.org/ just redirects to https://lumendatabase.org/... So while there's an addition of "Lumen Database" text, I suspect that should be the link target, not Chilling Effects, and update to avoid the redirect...

So <ins>Lumen Database</ins> [https://www.chillingeffects.org/ Chilling Effects] website should become <ins>[https://lumendatabase.org/ Lumen Database]</ins><s>[https://www.chillingeffects.org/ Chilling Effects]</s> website. Potentially even dropping the website at the end too.

Reedy (talk) 00:06, 22 February 2023 (UTC)

Let me get back to you on this -- the copy for upload that we received did not strike the word "Chilling Effects" through. RamzyM (WMF) (talk) 00:52, 22 February 2023 (UTC)
Replying here. Lumen and Chilling Effects are the same organization, they just changed their name a few years ago. We had originally left it since they kept the old URL, but I think we can update to have it say Lumen and use their main link instead of the chilling effects link. -Jrogers (WMF) (talk) 00:57, 22 February 2023 (UTC)
Thanks, updated. RamzyM (WMF) (talk) 01:03, 22 February 2023 (UTC)
Could it be worded in a more change-proof manner? Maybe referring to some text law that designate this organization for example? Psychoslave (talk) 10:39, 28 February 2023 (UTC)

Day 1 thank you

Hi everyone, this is Jacob from the Foundation legal department. I want to say thank you for the initial comments! There were some good catches on the copy-editing, and helpful requests for adding and fixing links, which we'll review (including any more as they come in) and aim to produce an updated draft. I think we'll likely have that before the first office hour in March. I also reached out to the security team regarding the third party resources policy. Thank you Taavi and xaosflux. It looks like that is the correct phabricator task, but their policy isn't ready yet. We are likely going to keep this space in the ToU as a placeholder for the future when there will be a policy for how to use third-party resources without creating a technical risk for other users. -Jrogers (WMF) (talk) 00:55, 22 February 2023 (UTC)

Password Security

I don't think we should get into specifics of password policies, good practices like using password managers etc (which are technical) here, but is it worth going a little beyond the current wording of "You are responsible for safeguarding your own password and should never disclose it to any third party." to include some wording about changing it if you believe it's been shared/compromised. Reedy (talk) 01:42, 22 February 2023 (UTC)

Though, what is a "security credential"? Does that need defining or examples providing? Reedy (talk) 01:44, 22 February 2023 (UTC)
Probably covers things like 2FA codes. Dreamy Jazz talk to me | enwiki 12:35, 23 February 2023 (UTC)
The longer the ToU is the less likely it is that people are going to read it. Why do you think it's valuable to add more words? What effect do you think the additional words will have? ChristianKl ❪✉❫ 14:07, 22 February 2023 (UTC)
I concur with Christian: Keep it short. However, I am sort of missing a frame of reference with respect to the current state of the art. So I agree with you, Reedy, too. Using weak or even compromised encryption for your password manager should constitute inadequate/insufficient password security. ‑‑Kays (talk) 23:38, 24 February 2023 (UTC)
The point of this line in the ToU is to remove the excuse of "I didn't do it, somebody else did it by using my password"-excuse in a legal context. The existing words do that. It's not to inform the user about the best way to keep their password safe. ChristianKl ❪✉❫ 13:46, 26 February 2023 (UTC)
If it is meant to mean something like "your may be held responsible for abuse of your password by someone else", then I think it should say that explicitly. PJTraill (talk) 14:42, 4 March 2023 (UTC)
Short and simple: "It is your responsibility to protect your Login." or some such. (This avoids reference to the exact technical methods—eg. SSH keys; thought doubtful anything is needed at all). —Sladen (talk) 17:05, 20 April 2023 (UTC)

Responsibility for laws where subjects written about are located?

The Overview section currently states, "you are legally responsible for all of your contributions, edits, and re-use of Wikimedia content under the laws of the United States of America and other applicable laws (which may include the laws where you live or where you view or edit content)." The proposal changes this to, "you are legally responsible for all of your contributions, edits, and re-use of Wikimedia content under the laws of the United States of America and other applicable laws (which may include the laws where you or the subject you’re writing about are located)." [Emphasis added.] This is a very substantial and disturbing change. Will editors writing about the Russian military be bound to follow the law of Russia, for example? Is the Foundation requiring editors to learn and follow the libel laws of foreign countries of people and corporations? And if they run afoul of those laws, does this proposed revision make editors liable for the consequences? The "About" page's Overview section says, "These changes are open to general feedback and can be removed or reworded if there are concerns." Yes, please! 2601:642:4C02:64B:CE6C:2AA:1BDB:EF2 05:07, 22 February 2023 (UTC)

i also share this concern. Bawolff (talk) 08:25, 22 February 2023 (UTC)
@Bawolff I'm not sure this guidance actually says anything unless it is an indication that WMF is going to enforce something themselves. It seems that it is just a warning to contributors that if you publish something it is on you, and some governments may not like it and go after you. — xaosflux Talk 12:01, 22 February 2023 (UTC)
The line that stood out to me was in the summary "Under the following conditions:...Lawful Behavior — You do not violate copyright, post illegal content or violate other laws.". I mean, I imagine the intention here is don't post CP type illegal content, but i can't help but think it implies that we are frowning on someone using a VPN to edit from china. Bawolff (talk) 12:20, 22 February 2023 (UTC)
Historically, one of the tasks of the Wikimedia legal department was to defend Wikipedia editors when they are sued for the good intentioned work they do on Wikipedia. Using wording like this is what the can use to say "You are on your own, it's not our problem that you got sued". ChristianKl ❪✉❫ 13:35, 22 February 2023 (UTC)
The current version of the ToU (which went in effect 2014) has a similar wording, with exactly the meaning Xaosflux suggested. This wording doesn't mean „it's not our problem“, the WMF might of course offer help, but it's very important to warn editors about potential legal risks when living in or writing about authoritarian regimes. Johannnes89 (talk) 13:50, 22 February 2023 (UTC)
To confirm, we added this to help editors know to be careful when working on BLP articles. It won't limit the Foundation's legal fees assistance or cause the Foundation to take any action against editors. We've seen a number of courts around the world over the past decade find jurisdiction based on the location of the subject of an article rather than the editor's location and we thought people should know that's a risk. -Jrogers (WMF) (talk) 19:22, 22 February 2023 (UTC)
@Johannnes89 and Jrogers (WMF): Does WMF have a legal obligation to warn users? I’m cautiously questioning whether the ToU are the right location for a mere “courtesy warning”. ‑‑Kays (talk) 23:39, 24 February 2023 (UTC)
I'm not a lawyer, and am a relatively new editor, but the text in section 1 (immediately following the section quoted) says "Although we may not agree with such actions, we warn editors and contributors that authorities may seek to apply other country laws to you", which seems pretty important to this point. Skyvine (talk) 23:24, 22 February 2023 (UTC)
Whatever the Foundation "may not agree", I understand it could agree in some case, or not agree and accept any legal attempt. I share the two concerns (first one and ChristianKl's one - "You are on your own, it's not our problem that you got sued"). --Pa2chant.bis (talk) 06:11, 23 February 2023 (UTC)

(Section OP here.) Could we please clarify that violating the laws of a country other than those of the US or where the editor is located is not a ToS violation, and that the Foundation will resist efforts to prosecute such accusations? 2601:647:5701:39B0:B973:DF04:7A6D:ADCA 16:16, 13 March 2023 (UTC)

Third party resources

The new section about "Unauthorized Uploading of Third-Party Technical Resources" is very unclear to me. Like, i am a developer-definitely more technical than most, and i have no idea what that is trying to get at. Are we trying to ban people from reusing open source js libraries in gadgets? Installing rootkits on the servers? Something else? Bawolff (talk) 08:22, 22 February 2023 (UTC)

Hello @Bawolff:, please see the comment I made in Taavi's section a bit earlier: Special:Diff/24608405. Thank you. -- Samuel (WMF) (talk) 11:35, 22 February 2023 (UTC)

Sui generis database rights

One of the key changes in the proposed Terms of Use is the new sentence about contributors waiving their "Sui Generis Database Rights". This is one of the key issues that delayed the move from 3.0 to 4.0 for so long (see answer by Jimmy Wales in a 2021 conversation with the WMF Board of Trustees).

The background here is that Wikidata has a CC-0 (public domain) licence that does not require re-users to attribute the content to Wikidata, whereas re-users of Wikipedia do have to attribute content to Wikipedia. Because of this licence incompatibility, it was originally thought that it would be a licence infringement for Wikidata to harvest data from Wikipedia, as has been happening for the past ten years. (In fact, because of the licence incompatibility, the community were told in 2012 that no such harvesting would ever take place.) The licence incompatibility would have been even more of an issue under the 4.0 licence, which explicitly protects database rights. As the Creative Commons update for Version 4.0 said:

In particular, the fact that sui generis database rights are not explicitly covered by the 3.0 unported licenses has led to confusion in jurisdictions that recognize those rights. Version 4.0 removes any doubt, pulling applicable sui generis rights squarely within the scope of the license unless explicitly excluded by the licensor. It also allows database providers to use the CC licenses to explicitly license those rights.

Per the video linked above, the decision to go ahead with the move from 3.0 to 4.0 now is in part due to the UN's decision to adopt 4.0. Andreas JN466 11:38, 22 February 2023 (UTC)

Per the video linked above, the decision to go ahead with the move from 3.0 to 4.0 now is in part due to the UN's decision to adopt 4.0. - that's not how I understand that part of the video. It was a response to a questioner who brought up a *potential* decision of the UN to adopt 4.0 for its text content - has that actually happened in the meantime? - alongside the fact that lots of academic journals have already do so. I read the response as Jimmy, in his usual affable way, agreeing with the questioner that that might be an argument for speeding up (or reviving) a process that had already been set in motion earlier based on the positive community feedback in 2016 ("this is clearly the direction of travel"). But your causality claim seems unfounded.
Personally I think that the second part, about open access academic journals, is much more important. I regularly encounter peer-reviewed research papers with 4.0 licensed text (on arXiv and elsewhere) that I think would be useful for Wikipedia (especially from literature review parts where authors are summarizing other people's research), and am very much looking forward to the ToU update enabling us to reuse them.
Regards, HaeB (talk) 08:04, 25 February 2023 (UTC)
Listening to it again, I think you're right.
I'm not au fait with the developments at the UN that the questioner was referring to. But the United Nations University for example certainly does release a lot of material under the 4.0 license, as do the many other academic journals you mention. Andreas JN466 09:49, 25 February 2023 (UTC)
And regarding Wikidata and "harvesting", you may want to define that term more precisely, in particular to substantiate your claim that it was originally thought that it would be a licence infringement for Wikidata to harvest data from Wikipedia, as has been happening for the past ten years. (In fact, because of the licence incompatibility, the community were told in 2012 that no such harvesting would ever take place.) - Denny may be able to clarify his 2012 comment that you are citing here (although he's not a lawyer either AFAIK). But my guess is that by "Wikidata does not plan to extract content out of Wikipedia at all" he was referring to copyrighted content from individual Wikipedia articles. In any case he can't have meant the large-scale creation of Wikidata items for all Wikipedia articles, since that had already been a core aspect of the plan for Wikidata at that time. - Either way, even if your theory were correct that the Wikidata project has been a huge violation of Wikipedia's database rights all along for the past decade because of the limitations of the 3.0 license (and I doubt that this is true), that would be a very strong argument for adopting the 4.0 license now, to avoid having to shut down Wikidata, which (warts and all) has been a successful and valuable addition to the Wikimedia movement. (I'm also confused why you claim that the 4.0 licence [...] explicitly protects database rights. That's not what the quoted text from Creative Commons means.)
Regards, HaeB (talk) 08:50, 25 February 2023 (UTC)
If what you say were true, the WMF would not ask us to accept the addition of a new sentence about waiving database rights:
Where you own Sui Generis Database Rights covered by CC BY-SA 4.0, you waive these rights. As an example, this means facts you contribute to the projects may be reused freely without attribution.
What Denny said ten years ago was consistent with WMF Legal's assessment, which is copied and linked below, in a section in which you commented (#Creative Commons imports and sui-generis database rights): The sui generis right is infringed when "all or a substantial part of the database" is transferred into another document or database. ... the Directive also prohibits the "repeated and systematic extraction" of "insubstantial parts of the contents of the database".
As I read it, it is only because 4.0 now pulls database rights (which were "not explicitly covered by the 3.0 unported licenses") "squarely within the scope of the license" that the WMF wants us to accept the above sentence about waiving database rights.
Incidentally, I believe the sentence the WMF aims to add to the ToU means two things:
  1. Wikimedia projects would not be able to harvest 4.0 databases.
  2. If any government body or other organisation were to add the content of their 4.0 database to a Wikimedia project they would thereby automatically give up their sui generis database rights, including the right to attribution. That is an end run around the actual intent of the 4.0 license, and a bullying move.
I don't like that. The Oxford Internet Institute's Mark Graham and Heather Ford have written eloquently about what it means when people have no way of telling any more where the data they are presented with actually comes from. We should resist rather than encourage this trend.
It's a problem that's presently getting far worse, by the way, courtesy of AI tools like ChatGPT – another potential customer sector currently being eyed by Wikimedia Enterprise. Regards, Andreas JN466 09:37, 25 February 2023 (UTC)
You haven't answered my suggestion above to clarify what you mean by (the rather loaded term) "harvesting". Do you think that Wikidata has been (or is) in violation of database rights for Wikipedia, or not? If yes, what precisely is the "substantial part" being "harvested"? Also in that case, would you really prefer having to shut down Wikidata for legal reasons over going through with the proposed ToU update?
If what you say were true - I said several things above, can you be more specific about which of these you think may not be true? ... the WMF would not ask us to accept the addition of a new sentence about waiving database rights - my guess is that this is to ensure that the default version of the 4.0 license will be consistently applied (i.e. without individual contributors applying the "unless explicitly excluded by the licensor" exceptions on a case by case basis in their edit summary or such). I don't see it as a tacit admission that Wikidata would need to be shut down for legal reasons otherwise. Also, I don't understand how the proposed ToU change would mean that Wikimedia projects would not be able to harvest [any] 4.0 databases, apart from those that have applied that exception and "explicitly excluded" some rights from the 4.0 license? By the way, while I'm not a Wikibase expert, it seems to me that tracking and maintaining such extra database-rights based requirements on content imported to Wikidata might be very cumbersome and onerous anyway, requiring substantial technical investments at least. In any case, I entirely disagree with your end run and bullying move characterizations.
Furthermore, considering that individual facts are not copyrightable (also not in the EU), it seems clear that the "facts you contribute" part would not affect the reuse of facts gleaned from Wikipedia (my emphasis) that you talk about below, anyway (only possibly database-like aggregations of such facts); that's what my comment there was about. Apropos, while I'm a huge fan of citations and attribution and en:WP:BURDEN myself, I have doubts about the existence of this alleged "trend" that people have no way of telling any more where the data they are presented with actually comes from - as if every newspaper article, book or TV documentary came with footnotes in the past.
Lastly, a major problem with your way of thinking about these topics is that you appear to constantly imagine reusers of our content as evil capitalist tech predators (AI! ChatGPT! or whatever other bogeymen there may be). But making dumps of our entire projects freely available without restrictions beyond the CC license has been a core practice since Wikipedia's very early years. Contrary to your insinuation, dumps aren't something that was recently started for the corporate customers of Wikimedia Enterprise. This ability to copy and reuse the entirety of a Wikimedia project's content has enabled countless mirrors and forks by amateurs, research projects by academics, etc. Preserving this right to fork is also very important for the community itself, to limit the power of WMF, something which I think you care deeply about. (For example, the famous 2002 Spanish fork has been claimed as a major reason why Wikipedia doesn't have ads today. Conversely, the absence of database restrictions - at least in the US - also enabled the Wikitravel community to fork the entirety of their CC-licensed project content after the site had been sold to a rather exploitative commercial wiki provider owned by a private equity investor, to reestablish it as a Wikimedia project, and to prevail in the ensuing lawsuits.)
Regards, HaeB (talk) 12:26, 25 February 2023 (UTC)
As I understand it, under 3.0 there was "wiggle room". This was acknowledged by Creative Commons in their "What's New in 4.0" release notes – "the fact that sui generis database rights are not explicitly covered by the 3.0 unported licenses has led to confusion in jurisdictions that recognize those rights" is how they put it. This uncertainty was also reflected in WMF Legal's m:Wikilegal/Database_Rights#Conclusion (my emphases): It can be difficult to determine whether, or the extent to which, a database or its contents are protected by law. Concepts like creative choice and substantiality are inherently subjective and it is rarely easy to predict how a court might rule under either US copyright law or the EU Database Directive.
Note, however, that they continued (again, my emphases): Whenever possible, the best course is to use only content that is made available by the author under an open license. In particular, for EEA and UK databases, the license should include a license or express waiver of the sui generis database right. In the absence of a license, copying all or a substantial part of a protected database should be avoided. Extraction and use of data should be kept to a minimum and limited to unprotected material, such as uncopyrightable facts and short phrases, rather than extensive text. For EEA and UK databases, bots or other automated ways of extracting data should also be avoided because of the Directive’s prohibition on “repeated and systematic extraction” of even insubstantial amounts of data.
This speaks to what I said earlier about importing data from 4.0 databases. It was also exactly the point that was at issue back in 2012 in that licensing discussion. The German user, User:Alexrk2 (who hasn't edited since 2021) said: CC0 won't be compatible with other projects like OpenStreetMap or Wikipedia. This means a CC0-WikiData won't be allowed to import content from Wikipedia, OpenStreetMap or any other share-alike data source. The worst case IMO would be if WikiData extracts content out of Wikipedia and release it as CC0. Under EU law this would be illegal. As a contributor in DE Wikipedia I would feel like being expropriated somehow. This is not acceptable!
It was in response to that comment that Denny said, Alexrk2, it is absolutely true that Wikidata under CC0 would not be allowed to import content from a Share-Alike data source. Wikidata does not plan to extract content out of Wikipedia at all. Wikidata will provide data that can be reused in the Wikipedias. And a CC0 source can be used by a Share-Alike project, be it either Wikipedia or OSM. But not the other way around. Do we agree on this understanding?
As I mentioned years ago in The Signpost (see Whither Wikidata?), this understanding then appears to have changed. A decision was made that there was enough ambiguity in the legal situation to risk performing large-scale, systematic imports of data from Wikipedia, thereby potentially infringing the Wikipedia community's database rights. After all, how likely was it that the Wikipedia community would sue Wikidata? Not very likely.
You ask me what I would like to see done with Wikidata. My preference has always been that Wikidata should have the same licence as Wikipedia, the main reason being that there should be traceability of provenance. Freebase had a CC attribution licence, so why not Wikidata? (Note that Freebase too was hoovered up by Wikidata and converted to CC0 ...) We are currently subordinating everything to tech companies' convenience and bottom line. Why?
The situation we're moving to is a gargantuan, hugely profitable data soup dished out by Big Tech via voice assistants and chatbots as "The Truth", without the public having the ability to trace and verify where these data actually come from. That does not empower the public – on the contrary. I couldn't find a review of Mark Graham and Heather Ford's 2016 paper in your Recent Research rubric in the Signpost; perhaps you missed it at the time (pity, I would have loved to see what you thought about it then ). It was preceded by a Slate article; have a look and see whether any points they make there about the importance of provenance ring true. (I wrote about that aspect too, years ago in the Signpost in the context of the Knowledge Engine, see "The times, they are a-changing".)
Lastly, you mention that pre-Internet newspaper articles etc. didn't come with footnotes either. The difference is that – even with all the problems around newspaper ownership – no newspaper had a similar market dominance to Big Tech. People always had a choice of Telegraph or Guardian, taz or Welt or what have you. As I see it, Big Tech is asking us to trust them to deliver the full breadth of opinion to us, neutrally – but even if they did that today, I wouldn't trust anybody to do that long-term.
People are human, which is why people have tried to build systems with checks and balances. That system is undermined if a lot of power ends up concentrated in a handful of trillion-dollar companies that then use their wealth to influence global politics and public opinion in a way that benefits their bottom line. (That is just as true with data as it's been true for oil.) To my mind, requiring answer machines of all kinds to say where they get their data from would make a small but important difference. It would make tech companies more accountable and transparent and reduce the likelihood of problems down the road. Andreas JN466 18:07, 25 February 2023 (UTC)
I sense some fundamental misunderstanding here: As I understand it, under 3.0 there was "wiggle room". This was acknowledged by Creative Commons in their "What's New in 4.0" release notes – "the fact that sui generis database rights are not explicitly covered by the 3.0 unported licenses has led to confusion in jurisdictions that recognize those rights" is how they put it. - It sounds like you think that 4.0 is stricter (grants less rights to reusers) than 3.0 in that regard, i.e. removes "wiggle room" because the 4.0 licence [...] explicitly protects database rights. But the quoted part refers to the opposite: Under 3.0, database rights were (at least according to some interpretations) not covered by the license, meaning that no permission was granted for them. That 4.0 is "pulling applicable sui generis rights squarely within the scope of the license" means that permission is now explicitly granted for them. (Remember, a license is basically a permission to do something that would otherwise be disallowed by existing rights - but never a tool to impose additional restrictions. Something being under the scope of a license means that it is covered by the permissions granted by the license.) This becomes clearer when looking at the full context of your quote from https://creativecommons.org/version4/ (see also their explanatory links within the text there, left out in the quote below):

Rights outside the scope of copyright
Other rights beyond copyright can complicate the reuse of CC-licensed material. To the extent that those rights are not addressed directly in a copyright license, the situation for users of works can be even more confusing. Version 4.0 addresses this challenge through an open-ended but carefully tailored license grant that identifies categories of rights that could (if not licensed) interfere with reuse of the material. Accounting for these and other unenumerated rights will more fully enable users of CC-licensed works to use the work as they expect and as intended by licensors.
In particular, the fact that database rights are not explicitly covered by the 3.0 unported licenses has led to confusion in jurisdictions that recognize those rights. Version 4.0 removes any doubt, pulling applicable sui generis rights squarely within the scope of the license unless explicitly excluded by the licensor.

Also, regarding This uncertainty was also reflected in WMF Legal's m:Wikilegal/Database_Rights#Conclusion : That Wikilegal page talks about an entirely different kind of uncertainty in that regard, namely what uses are covered by database rights - not whether the permissions granted by CC licenses cover these database rights (the subject of the confusion that Creative Commons' explanations refer to).
Regarding Wikidata, you still haven't responded to my question what precisely is the "substantial part" being "harvested", or in the wording of your most recent comment, which large-scale, systematic imports of data from Wikipedia you are referring to concretely. This makes it difficult to discuss how much they might really be affected by legal restrictions, and what the benefits and drawbacks of the proposed ToU update might be in that regard. Again, assuming that Denny was reasonably sober at the time of his 2012 comment, with "Wikidata does not plan to extract content out of Wikipedia at all" he can't possibly have meant that Wikidata would not import the entire list of the (around that time) tens of millions of Wikipedia articles (without text content but with interwiki links), because that was already a core part of the plan for Wikidata at that time, as a central first step to initialize its database, so to speak. Please don't keep simply repasting your favorite quotes (as you are wont to do), especially if it has already been pointed out that they may be taken out of context.
As for your thoughts on Big Tech, news media and the articles by Heather and Mark that (via the cited Signpost op-ed) amplified the theses from your Signpost piece, I think we're developing some off-topic tangents here (my fault too). But I'll be happy to respond a bit later because I do agree that these are important and interesting topics in general. I'm not here to defend "Big Tech" or to dismiss your concerns entirely, but I do think that some frequently proposed arguments are questionable, and partly based on flawed assumptions.
Regards, HaeB (talk) 03:30, 26 February 2023 (UTC)
PS: Concerning your fundamental misunderstanding about 3.0 providing "wiggle room" and the 4.0 licence [...] explicitly protects database rights, what I said above was actually already explained by the legal team in 2016: m:Terms_of_use/Creative_Commons_4.0/Legal_note#Waiving_database_rights (The update in the 4.0 version of the license additionally provides permission to use material that is not eligible for copyright protection, but is eligible for protection as part of a database, etc.). Regards, HaeB (talk) 03:54, 26 February 2023 (UTC)
My head hurts. How then would you interpret the following passages from https://wiki.creativecommons.org/wiki/data (my emphases):
How does the treatment of sui generis database rights vary in prior versions of CC licenses?
[...] In the CC version 3.0 licenses, the legal treatment of sui generis database rights varies, but the practical result is always the same: compliance with the license restrictions and conditions is not required where sui generis database rights - but not copyright - are implicated. This means that if a substantial portion of a CC-licensed database is extracted and used in a way that does not implicate copyright (e.g., by rearranging purely factual data), the license does not require the user to attribute the licensor or comply with any other restrictions or conditions, even if the database is protected by sui generis database rights. [...]
Cf. the last section on that page:
If my use of a database is restricted by sui generis database rights, how do I comply with the license?
If the database is released under the current version (4.0) of CC licenses, you must attribute the licensor if you share a substantial portion of the database contents. Andreas JN466 23:46, 26 February 2023 (UTC)
Well, we've been talking about the unported licenses in this thread ever since your first comment (which makes sense, considering that that's what the Wikimedia projects are using). Regarding your last quotes from https://wiki.creativecommons.org/wiki/data , as every so often, context matters. To repeat with the last "[...]" expanded and emphases changed:
How does the treatment of sui generis database rights vary in prior versions of CC licenses?
[...] In the CC version 3.0 licenses, the legal treatment of sui generis database rights varies, but the practical result is always the same: compliance with the license restrictions and conditions is not required where sui generis database rights - but not copyright - are implicated. This means that if a substantial portion of a CC-licensed database is extracted and used in a way that does not implicate copyright (e.g., by rearranging purely factual data), the license does not require the user to attribute the licensor or comply with any other restrictions or conditions, even if the database is protected by sui generis database rights.
While this result is the same across all CC version 3.0 licenses, the reason for this outcome varies. In the 3.0 licenses ported to the laws of EU jurisdictions, the scope of the licenses expressly covers databases subject to copyright and/or sui generis database rights. The conditions of the license are explicitly waived when use of the licensed work only involves the exercise of database rights.
By contrast, the 3.0 unported licenses and all other ported licenses do not expressly license sui generis database rights. As a result, those licenses do not apply when sui generis database rights alone are implicated. This means a licensee may need separate permission to use the database in a way that implicates sui generis database rights (although arguably an implied license to exercise those rights may be deemed granted in some jurisdictions). [...]
So in other words, yes, Wikipedia's current 3.0 license does not require attribution for these kind of "substantial" extractions - but it does not provide permission for them in the first place ;-) (at least not explicitly)
In related news, I do not require you to fill up the gas tank after you drive my car. But I haven't permitted you to drive my car. (Also, I don't own a car.)
Regards, HaeB (talk) 03:10, 27 February 2023 (UTC)
Well, forgive me, but that is precisely the point, isn't it? The Creative Commons licences were created to make it easy to define re-users' duties. They were created so commons creators would have an easy, widely understood and enforceable way to say, You are welcome to re-use our stuff – as long as you do x, y and z (for example, attribute, i.e. credit us).
We both agree that 4.0 expands the scope of CC-license-defined attribution requirements to cases where only sui generis database rights (based on the "sweat of the brow" argument) are involved. Right? We agree 3.0 didn't say anything about attribution in such cases. This means that 3.0 was of little help to anyone wanting to ask for attribution in cases where only sui generis database rights were involved. (And I think I am correct in saying that there is no universally agreed, easy-to-understand, well-tested alternative legal framework in place that would enable people to demand attribution based on sui-generis database rights: that is precisely the gap the CC licences are meant to fill.)
The example of your car is somewhat misleading because people generally don't want others to drive their car. This is not the situation we have here.
If the car is Wikipedia, or Wiktionary, etc., we want people to drive our car, because we made it for other people to use. But the CC BY-SA licence expresses our wish that people should (1) credit us while doing so (say where they got the car they're driving) and (2) share alike (leave the car somewhere others can pick it up when they're done with it, so they can use it as well).
If I had to come up with an analogy involving your car (it's difficult ...), I'd say that under 3.0, people only had to say "I got this car from HaeB" while using it to drive to work; under 4.0, they also have to say "I got this car from HaeB" when they repaint it yellow, stick their own name on it and use it to provide paid taxi services.
At any rate, adopting the 4.0 licence would increase our CC-based rights to attribution. The CC BY-SA 4.0 licence expands re-users' attribution and share-alike duties defined in the licence to another, previously uncovered set of circumstances (sweat of the brow).
And that is precisely the increase in rights the WMF wants us to agree to waive. Well, I don't agree. Regards, Andreas JN466 09:43, 27 February 2023 (UTC)
Andreas, you are continuing your fundamental misunderstanding. To repeat from the part of https://wiki.creativecommons.org/wiki/data I bolded above, under 3.0 unported a licensee may need separate permission for this kind of reuse - i.e. it remains prohibited by applicable database rights laws. In contrast, 4.0 permits it (the current version of the CC license suite (4.0) licenses sui generis database rights in addition to copyright and other closely related rights), with attribution requirements. You are only focusing on the attribution requirements for the reuse and are forgetting about whether the reuse is permitted (covered) by the license in the first place. So no, we do not agree that 4.0 expands the scope of CC-license-defined attribution requirements except in the utterly meaningless sense that it expands their scope to something that wasn't permitted by the license before anyway.
Hence it is wrong to claim that adopting the 4.0 licence would increase our CC-based rights to attribution. Since 3.0 unported doesn't cover these reuses, we could currently already demand attribution for them in any form we want (say in oscillating multicolored letters that include the phrase "ALL GLORY TO THE WIKIPEDIA", or alternatively ask for a donation of 10 trillion USD to the Wikimedia Endowment), when negotiating that aforementioned separate permission. 4.0 reduces that ability to demanding attribution in the very confined form specified by the BY licenses, which still may not even be practical for other reasons as per the legal team's rationale for this proposed ToU change (which you appear to have been ignoring entirely so far). Regards, HaeB (talk) 06:13, 28 February 2023 (UTC)
You say So no, we do not agree that "4.0 expands the scope of CC-license-defined attribution requirements" except in the utterly meaningless sense that it expands their scope to something that wasn't permitted by the license before anyway. Firstly, who is "we"? Secondly, and more importantly, I don't agree that this case where 4.0 does expand our CC-based rights is "meaningless". You argue as though people always asked for permission before putting stuff online. That's not how the Internet and the related case law have grown (Luis' 2016 blog post actually touched upon this). 4.0 is a powerful lever to help build an Internet where end users can see where information originated. The WMF wants us to give up our right to use it. Andreas JN466 08:08, 28 February 2023 (UTC)
Firstly, who is "we"? - umm, "we" was referring your own comment that I was replying to (you had claimed that We both agree that 4.0 expands the scope of CC-license-defined attribution requirements ... and I responded that we ( = you and I) actually do not agree, because of your fundamental misunderstandings about these legal texts). Sorry to call this out, but the fact that you are badgering me here about this word, with smiley template and all, because you misunderstood and forgot the context from your own preceding comment is rather illustrative of how you continue to misunderstand the legal texts and leave out important context.
I don't agree that this case where 4.0 does expand our CC-based rights is "meaningless" - see, that is the problem. Above I explained in detail why it does not "expand our CC-based rights", per [3]. (I assume the first person plural refers to the owners of sui generis database rights, which I suspect most Wikipedia editors aren't. In any case, yes, under 4.0 these database rights holders retain the right to demand attribution in some jurisdictions, but under 3.0 unported they have the same right and more. So quite the opposite of "expand".) Regards, HaeB (talk) 06:47, 6 March 2023 (UTC)
HaeB, point taken about the "we", apologies. (If you'd said "I do not agree with you that ..." your meaning would have been clearer.) The problem with the waiver clause is that it further normalises non-attribution, and not sharing alike. Under the 3.0 licences database rights were either not covered (unported licences) or waived (ported licences). But under 4.0 they are expressly included. 4.0 thus provides people with an easy way to insist that re-users attribute and share alike, based on database rights. That is a good thing in my view. But this is precisely what the WMF wants contributors to waive, so the pipeline from Wikipedia to Wikidata to Google, Apple, Amazon, OpenAI etc. can disgorge a maximum amount of data that does not require re-users to attribute and share alike.
To my mind, this is all part of a trend to arrange the data environment in such a way as to empower commercial players to do whatever suits their financial interests best – not building systems for attribution, because it is burdensome and would cost money, and not sharing alike, because proprietary and non-transparent products are more profitable than open and transparent ones. But as useful as it may be for promoting profits and commercial development, it also involves considerable downsides. It destroys the end user's ability to see where information comes from, and results in free content generated by volunteers being converted first into CC-0 and then into proprietary commercial content. Andreas JN466 14:47, 14 March 2023 (UTC)
Just to not stay silent: I'm firmly against such a change.
I would rather like to see the exact inverse move: to use this right as extensively as possible. Promote and defend the diffusion of free-libre-open-works (FLOW) through licenses which dispense inherited resistance.
But I don't think it would happen even if a majority of active contributors would come and say "I agree let's do that" following this message, so I won't waste my time further on this lost battle. To my mind, the forces causing this change are not community driven nor done for the benefit and defense of the community. Psychoslave (talk) 23:28, 27 February 2023 (UTC)
I don't think it is a "lost battle"; the WMF has already shown willingness to change parts of the proposed ToU update based on community feedback. However, if you want to make a convincing case against this particular change, you should really first read the legal team's rationale for it (including the blog posts linked there, by Luis, who long before and after his time as WMF's Deputy General Counsel has been known as a free/libre/open source advocate), and explain how you would handle the difficulties that they want to avoid with it. To be clear, I haven't gone through these rationales in detail myself either. (I'm mostly just jumping in here to point out some obvious misunderstandings.) But generally speaking, there are things that sound great in theory but in practice would create a mess of inconsistent and unworkable requirements on the one hand and do not even achieve the desired goal on the other hand (say if the particular reuses one worries about are covered by fair use anyway).
Also, what does licenses which dispense inherited resistance even mean? Regards, HaeB (talk) 06:50, 28 February 2023 (UTC)
I don't know Luis, but it is only to be expected that a lawyer with a significant history of work for Google and other tech companies will come up with arguments that are aligned with those companies' interests. No disrespect to him, but on matters such as this I am more interested in hearing the views of independent experts, people who are less used to seeing and presenting things just from tech companies' perspective. Andreas JN466 09:09, 28 February 2023 (UTC)
Also, what does licenses which dispense inherited resistance even mean?
A license that preserve copyleft on derivative works, such as GPL, CC-*-SA or ODBL. Psychoslave (talk) 09:31, 28 February 2023 (UTC)
That is an important point, because each time CC BY-SA rights are waived the result is that work done for the commons ends up being transformed into proprietary products. Andreas JN466 09:40, 28 February 2023 (UTC)
Looking at the rationale, I respectfully but strongly disagree with it.
What if we were saying "ok not every country have the same protection rules for employees, so let’s align with the lowest common denominator". It might be great for the evil-big-corp Leviathan which crave for the cheapest right-less possible workforce, steered as it is with only a narrow purely financial view of the world. But for evolution of societies at large, my opinion is that this is an awful line of conduct.
I wonder how the legal team would react if we would propose to align their own professional conditions to the combination of the worst professional equivalences in the world. 😏
I think we should strive for putting the bare as high as possible regarding respect of each human not at lowering it. In present context, preserving attribution and an equitable conservation of the same rights on derivative works would be the minimum fairness we should expect. Psychoslave (talk) 09:53, 28 February 2023 (UTC)
I don't think it is a "lost battle"; the WMF has already shown willingness to change parts of the proposed ToU update based on community feedback.
Well, I have no doubt that inputs like the ones I made how some sentences are formulated have their chances.
But for that kind of topic, I am so confident that I’m peeing into a violin, that I very willing to bet with you that nothing substantial will be change on this statement. So, tell me, what would you like to bet? 🙂 Psychoslave (talk) 10:33, 28 February 2023 (UTC)
Ok, I couldn't resist to waste a few more minutes and come with a divergent proposal:
Where you own Sui Generis Database Rights covered by CC BY-SA 4.0, you retains these rights to the full extent defined in the license. As an example, this applies to your contribution in novel data schemes as well as any contribution in original collect and curation of factual information, including in the form of cohesive data set. Thus guarantying that derivative works remains freely reusable, preserve attribution and help to improve data traceability.
Sure, I know this won't happen. But at least no one will be able to say that is just because no one in the community proposed something else. Psychoslave (talk) 23:51, 27 February 2023 (UTC)
No, that's not how this works, we can't use the ToU to expand sui generis database rights to anything we would like attribution for, as you attempt to do here: As an example, this applies to your contribution in novel data schemes as well as any contribution in original collect and curation of factual information, including in the form of cohesive data set. Thus guarantying that derivative works remains freely reusable, preserve attribution and help to improve data traceability. For example, I would assume that almost no manual Wikipedia edits are covered by these database rights, and also don't forget that they only apply in certain jurisdictions (e.g. not in the United States). Regards, HaeB (talk) 06:50, 28 February 2023 (UTC)
Keep in mind that I’m not a layer while reading the following answer. 🙏
No, that's not how this works, we can't use the ToU to expand sui generis database rights to anything we would like attribution for
Maybe this is misreading "retains rights to the full extent"? This is not intended to mean "extending beyond", but preserving completely what is already defined so far.
For example, I would assume that almost no manual Wikipedia edits are covered by these database rights
Maybe the hypothesis inducing this assumption is that database should be thought in term of informatics, like a SGBD, when here its actually taken here in a far broader legal sense. I would be extremely surprised that even subset of wikis like categories curated by the community doesn’t match Sui Generis clauses.
they only apply in certain jurisdictions (e.g. not in the United States).
So what? This is just a less appreciative way to say that in many jurisdictions, like all European countries where it does apply.
ToC should not be a place used to explicitly lower the the bar at every opportunity. Because, as we all know, the extend of US free speech is rather exceptionally high. I don’t see anyone pretending that we should lower the bar of expected speech rights that contributors possess just to align with those of European countries, let alone North Korea. Psychoslave (talk) 10:24, 28 February 2023 (UTC)

For the record, my preference would be to replace

  • Where you own Sui Generis Database Rights covered by CC BY-SA 4.0, you waive these rights. As an example, this means facts you contribute to the projects may be reused freely without attribution.

with the following wording:

  • Where you own Sui Generis Database Rights, you agree to license these rights under CC BY-SA 4.0.

Andreas JN466 14:47, 1 March 2023 (UTC)

Lawful Behavior

"Lawful Behavior — You do not violate copyright, post illegal content or violate other laws." this seems a bad idea. Totatlitarian states have all sorts of laws that limit free expression. I vaguely remember a case in the past where the Ugandian Wikipedia wanted to stop users from saying that they are homosexual on their user pages and the Wikimedia Foundation had a problem with that. Given that homosexuality is illegal in Uganda the current wording likely require such a policy from the Ugandian Wikipedia.

In the last month we saw Twitter censor links to BBC's Modi documentatry under the idea that it violates Indian law. A policy where Wikimedia would also censor that links seems like a violation of the values for which Wikimedia was created.

When totalitarian governments attack freedom of expression the role of the Wikimedia Legal Department shouldn't try to minimize legal exposure but to fight for freedom of expression. ChristianKl ❪✉❫ 13:32, 22 February 2023 (UTC)

An interesting point - it would also seem counter to the human rights policy, at least if read in its expansive form. Nosebagbear (talk) 17:01, 22 February 2023 (UTC)
This is a helpful comment, thank you for flagging it! The wording change in section 10 (which is what led to the updated summary this is quoting) is one of the requirements of the Digital Services Act, which requires a more broad unlawful behavior analysis than just copyright law. In practice, the risk of censorship here should be mitigated by a couple things. One is that the applicable law to the Foundation doesn't cover every country in the world, so problematic laws may not apply. Second is that the human rights policy should actually guide the Foundation's behavior here and the Foundation should not try to assist with the enforcement of a law that violates human rights if at all possible. I'll take this one back to review by the team to see if there's a way we can update the wording here to more clearly reflect the human rights aspect while still meeting the needs of the DSA. -Jrogers (WMF) (talk) 19:31, 22 February 2023 (UTC)
Also makes me wonder about the fate of the en:Illegal number article or mentions of Bloody Sunday's "Soldier F"; "illegal content" without specifying a jurisdiction or other context is really pretty broad. --20:01, 22 February 2023 (UTC) Dogcow (talk) 20:01, 22 February 2023 (UTC)
  • I also wanted to do illegal things like sharing general LGBT+ info and medical information of the sort which is conventional and would be taught in any Western medical school. As a general guide, the kinds of information in university research libraries is in the interest of Wikipedia, even when censored in some countries. It seems like WMF is on this though and the intended interpretation is to mean illegal content outside the scope of Wikipedia as presently understood. Bluerasberry (talk) 00:49, 23 February 2023 (UTC)

I suggest the following alternative:

  • Lawful Behavior — You do not violate copyright or other laws of the United States of America.

Long5563 (talk) 08:21, 23 February 2023 (UTC)

The WMF seems to want to write the policy in a way honors the Digital Services Act. To do that, it would require honoring EU law.
Generally, I think violations of the Official Secret Act of the UK should not be violations of the Wikimedia terms of use. Linking to BBC Modi documentatry as Wikipedia seems to do in violation of Indian law should be possible. Free speech is under attack in many parts of the world.
When it comes to LGBT+ info, violation of the Ghanian and Ugandians laws should be legal, but I expect that there are also other countries that have laws that forbid content we want to allow. If a country like Russia has a law that restrict information about LGBT+ we want the Russian Wikipedia to be still be able to provide that information.
Wikimedia should not only help users in the US or EU to write such illegal content in such jurisdictions but also give users in countries where their freedom of expression is limited a way to publish information.
The key jurisdiction in which the WMF operates is the United States, so requiring complying is United States law makes sense. Fortunately, the First Amendments is good at protecting speech. You seem to count EU law also as applicable law to the foundation. It's unclear to me which jurisdictions the foundation considers to have applicable law. The terms of use would be exactly the location where it would be worth to be specific so that users know what's allowed or not allowed in Wikimedia projects. ChristianKl ❪✉❫ 20:20, 23 February 2023 (UTC)
NLT and uncensored are the considerations for me. Long5563 (talk) 05:06, 24 February 2023 (UTC)
I agree that explicitly stating under which jurisdictions the WMF considers content to be illegal is paramount if this clause is accepted. Else, almost everything will be illegal — DaxServer (t · m · c) 08:56, 24 February 2023 (UTC)
@ChristianKl: May I remind you this is in the preamble. This summary is explicitly not legally binding. I think “don’t infringe laws” is a fair summary, like, generally speaking. We certainly don’t want to encourage people to break the law, right? ‑‑Kays (talk) 23:48, 24 February 2023 (UTC)
@Kays: I do want to encourage people in totalitarian countries to publish important information that's censored in their countries on Wikimedia. Providing ways for people in totalitarian countries to inform themselves about important issues is a valuable function that Wikipedia fulfills. Resistance against totalitarianism is about breaking the law and it's worthwhile to encourage.
Take the Ukraine war. Do you want to encourage Russians on the Russian Wikipedia to break the law and call it a war or do you want to discourage them from doing so? If they have a discussion about whether or not to allow that on the Russian Wikipedia I don't want the argument "The ToU says we don't violate laws" to be there. ChristianKl ❪✉❫ 02:06, 25 February 2023 (UTC)
I was also struck by this vague clause and agree with User:ChristianKl that it is probably especially problematic for people who come from countries with bad/archaic/problematic laws including those relating to blasphemy, nationalist symbolism, traditions, positions of power etc. "Whose law?" would definitely be a question that would arise. Shyamal (talk) 03:54, 25 February 2023 (UTC)
I would suggest this: Lawful Behavior — You do not violate United States copyright law, post content that is illegal under the law of the United States of America, or otherwise do anything that violates the laws of the United States.
The US's expansive understanding of free speech under the First Amendment is a major reason why Wikimedia is incorporated here. This would mean that it is not per-se a violation of the terms of service to violate foreign laws that suppress free expression.
--Rockstone35 (talk) 00:06, 26 February 2023 (UTC)
I also came to comment on this. It would be helpful to define concepts like "lawful" or "illegal" in the terms of which specific legal system. If left open to "any legal system" that leads to contradictions as various legal systems have laws that clearly contradict each other. Is there a way to clarify if, when using terms like "lawful" or "illegal" or "violate laws", the document means "of the United States" or "Of the jurisdiction you are located in" or whatever? Because one could read it either way. --Jayron32 (talk) 16:37, 1 March 2023 (UTC)
I am deeply concerned about this provision and would echo the perspective that others have shared above, that much of the content of Wikipedia would violate the laws of authoritarian regimes. Wikimedia shouldn't give such regimes legitimation in this way. Jmbranum (talk) 03:15, 2 March 2023 (UTC)
I agree that the current wording is much too vague. We should not be restricting our content to the lowest common denominator of what is considered legal, even in Europe. This would lead to substantial disruption, especially on Commons, which abides by U.S. law exclusively, except in the case of copyrights. Nosferattus (talk) 23:44, 2 March 2023 (UTC)

Agree the wording "illegal content" is confusing and undefined. If the terms of use are an enforceable document and may be produced to support or defend against legal action, then the policy document needs a definition that most readers will understand. if the wide interpretation discussed on this page applies, and website contributors suffer legal exposure, then the WMF have a duty of care to warn people before they edit about the nature of that risk rather than a casual caveat emptor approach which could seem callous, if not negligent, considering the consequences can include lengthy imprisonment. --Anstil (talk) 17:09, 3 March 2023 (UTC)

My eyebrows also went up at seeing 'unlawful' for most of the reasons above. Which laws users should stay within should be more explicitly stated. Given the dispute about online content laws in California, should it be 'Californian law' even if that means moving the WMF? 81.110.15.28 09:11, 4 March 2023 (UTC)

I want to +1 to this thread. The summary wording brought me to the talk page. Indeed I believe we have responsibility to allow people to write content against unjust laws. For the same reason we would consider changing the incorporation of WMF to another country, if it were to revoke its first amendment.--Thuvack (talk) 07:38, 5 March 2023 (UTC)

Agree. Change from "Lawful Behavior — You do not violate copyright, post illegal content or violate other laws." to "Lawful Behavior — You do not violate copyright or other relevant law." and avoid the word "USA" too. Taylor 49 (talk) 08:34, 5 March 2023 (UTC)
Agree. This change will help avoid any confusion about which jurisdiction is specified. FusionSub (talk) 10:06, 7 March 2023 (UTC)

Yes, this needs to be specified to "illegal in the United States". We shouldn't be following any other jurisdiction's laws, since the US does have the most expansive free speech guarantees in the world. Specifically, we should not be following any EU laws. Seraphimblade (talk) 02:09, 7 March 2023 (UTC)

Considering there are parts of the USA that are trying to put drag artists and parents of trans children in prison, just for "being", it's a potentially very narrow version of free speech, putting doubt on the long term future of US legislation being the one erratic horse to gamble everyone's future access to human knowledge on. --Anstil (talk) 09:24, 8 March 2023 (UTC)
@Anstil US laws aren't perfect but the Wikimedia Foundation has to follow them as long as the Wikimedia Foundation is headquarted in the United States. It might be useful to move the Wikimedia Foundation to Iceland to avoid being subject to US laws, but that would be a lot of effort. ChristianKl ❪✉❫ 17:32, 12 March 2023 (UTC)

Hi all, I wanted to come back to this discussion and flag that a very similar question came up in the office hours on March 2nd. We just got the transcript published so that's available if you want to review. I also want to point folks to our legal policies section on applicable law on meta which we updated in 2021 to talk about how we evaluate when a non-US law may apply to the Foundation and require our compliance. Also, I do want to note that we're working on an update to the draft under discussion this week based on the first few weeks of comments and we should have some changes based on this conversation to more clearly indicate that laws need to line up with human rights. Thanks again for the feedback and good discussion on this topic, I think it's a really important one! -Jrogers (WMF) (talk) 00:29, 15 March 2023 (UTC)

For the laws of other countries, we can draft a independent policy or guideline. But no policy may supersede the restrictions of United States law. Long5563 (talk) 13:01, 17 March 2023 (UTC)

Sticking to US laws may become problematic in the near future, see here: "Wikipedia has not explicitly said it could go under. But in a Supreme Court brief, it said it owes its existence to Section 230 and could be forced to compromise on its non-profit educational mission if it became liable for the writings of its millions of volunteer editors." Count Iblis (talk) 04:48, 17 March 2023 (UTC)

Age of Majority

Generally, it's good to keep policy documents short so that it's easier to read them. What changed in laws around the world that necessitates the change in wording about the age of majority? "Soliciting personally identifiable information from anyone under the age of 18, or under the age of majority where you are if higher than 18, for an illegal purpose or violating any applicable law regarding the health or well-being of minors." ChristianKl ❪✉❫ 13:39, 22 February 2023 (UTC)

To my knowledge, there is at least one state in the US (possibly two) where the age of majority is above 18. This does produce a weird situation though; does it violate the Terms of Use if you live in a jurisdiction where the age of majority is 18, but you're trying to solicit personal information from someone who is 18, but lives in a jurisdiction where the age of majority is older? --Rockstone35 (talk) 22:49, 22 February 2023 (UTC)
@Rockstone35: No, it’s referring to the potential solicitor’s age of majority in their current jurisdiction. ‑‑Kays (talk) 00:00, 25 February 2023 (UTC)

Marketing Company Mediation

Regarding the proposed terms of use update, what's the idea behind "Marketing Company Mediation"? The way UPE currently works is paid editors ignore the ToS, astute Wikipedians spot them and create a sockpuppet investigation, the UPE's current account gets blocked, the UPE makes a new account. Is there a situation where UPE would ever end up in court, and if so, why would mediation be superior to other processes? Thanks for your time. –Novem Linguae (talk) 15:33, 22 February 2023 (UTC)

I'm also confused. One interpretation is that the WMF seems to think it can compel anyone it suspects of UPE to travel to California. Although I have never carried out paid editing (declared or not), and am confident that any extradition hearing you might request would be laughed out of court, I would not wish to continue editing Wikipedia under those circumstances. Please clarify. Certes (talk) 19:33, 22 February 2023 (UTC)
Hi. I am also a bit confused about what Marketing Company Mediation is all about. — Jules* talk 19:45, 22 February 2023 (UTC)
Good questions. I'll try to be as simple as possible, but feel free to ask for more details.
1. The way UPE currently works is paid editors ignore the ToS, astute Wikipedians spot them and create a sockpuppet investigation. Yes it's true that editors diligently spot and rectify a large and meaningful amount of UPE, so that wouldn't change. But even in this best-case scenario, bad actors are only found out "after the fact." They've already created headaches for the good-faith editors who need to identify them, revert their edits, and potentially debate with these bad actors. This new process won't change the need for that, but to help relieve some of the burden and heartache that editors cleaning up this mess to deal with, the Foundation is asking for better tools to -- in parallel -- take affirmative action against the largest-scale bad actors.
There are -- for example -- large marketing firms that make hundreds of thousands of dollars every year just from charging people to create or edit Wikipedia pages. Some are outright scams (for example, guaranteeing pages to anyone and then taking their money). Some are much more sophisticated: using dozens of socks, using otherwise "clean" sleeper accounts, hiring sub-contractors to attack good faith editors. Some are even responsible for international corporate disinformation campaigns.
2. If so, why would mediation be superior to other processes? Mediation is superior to litigation for several reasons. Litigation is expensive and time-consuming, and the Foundation has not filed litigation in the past despite high-profile UPE controversies on the projects. This is not necessary because it lacks the interest to: for example, the en-Wiki community once overwhelmingly requested that the Foundation help take action against a particular bad UPE actor in the past. We were not able to act upon it because of the huge amount of time and cost barriers.
Mediation, on the other hand, would be much less costly and time-consuming than litigation. It would also require bad actors to reimburse the Foundation for the cost of their actions and legal expenses related to taking them to mediation, further lowering enforcement costs. The outcome of the mediation could be legally enforced in their jurisdiction as a court order if the entity does not comply. Overall, the mediation process could lead to an outcome with a fraction of the cost and time it would take for litigation.
3. One interpretation is that the WMF seems to think it can compel anyone it suspects of UPE to travel to California. There would be no traveling involved: mediations would be teleconferences according to the terms as currently written. In terms of jurisdiction, the marketing company would be subject to the same laws that govern contracts in that jurisdiction. Said differently, if the company is lawfully bound to the TOU, then they'd be lawfully bound to comply with the dispute resolution terms in the TOU should they create a dispute.
Should a marketing company believe that they are not bound, then that would be a matter for a court to decide in that relevant jurisdiction. We would attempt to get a court order in the company's jurisdiction, asking the court to compel the entity to show up to video-conference mediation (in any country of their choosing). This is a standard practice to enforce, so although we cannot predict it would be 100% effective in every circumstance in every jurisdiction, it's an effective practice with lots of precedents in international contract law. And even a failed court order would be magnitudes cheaper than litigation. There'd be no sort of "extradition" involved.
I have to head to a meeting, so sorry for the truncated reply, but I hope that addresses some of the questions at a big-picture level. SSpalding (WMF) (talk) 19:59, 22 February 2023 (UTC)
Question: response to the general question: what Marketing Company Mediation is all about.
The Foundation hopes to strengthen its tools to support existing community policies against marketing companies engaged in systematic, undisclosed paid editing campaigns. A "Marketing Company Mediation" would require a company providing services the community has identified as rule-violating UPE to submit to binding mediation to resolve the issue. The outcome of the mediation could be legally enforced in the company's jurisdiction as a court order should the entity not comply.
Even without lots of enforcement from the Foundation, we believe the language update alone would force many marketing companies into greater compliance with on-project rules. Rules against UPE already exist on the projects and in the TOU itself. But companies that make six figures a year providing these services have lawyers who may counsel them that it's worth the risk of a lawsuit (under the current difficult and costly enforcement process). Making it much cheaper for us and much more costly for the bad actors (should they lose) may change their risk tolerance for breaking project rules.
One of the first actions we'd likely take (if this TOU addition is seen as useful to the community) would be to send notices of the companies engaged in the most egregious scams telling them about this new enforcement mechanism. We would let them know that we know what they are doing and would be ready to enforce should these activities continue. SSpalding (WMF) (talk) 21:18, 22 February 2023 (UTC)
Thanks for the (crystal clear) answers! Those proposed changes seem welcome, as UPE are a pain. — Jules* talk 21:38, 22 February 2023 (UTC)
There's a part of me that worries that this could inadvertently benefit companies at the expense of volunteers and our broader commitments to accuracy and RS. If specific language for a Wikipedia article is agreed to in an arbitrated contract, do we have to put a notice for editors on the article "Don't touch this language at all or risk ToU violation"? What if new reliable sources suggest the agreed upon info is outdated, do we have to keep it pending a new agreement? Will the arbitrator have sufficient knowledge of how Wikipedia works to be a good guide for both parties? All I can say is I hope the WMF legal team will know what it's doing and not basically turn UPE into corporate pay to play. -Indy beetle (talk) 22:34, 22 February 2023 (UTC)
I'd like to +1 Indy's query. @SSpalding (WMF): - does this mediation/arbitration risk having content for any of the articles those UPEs working on being included within the outcomes? US arbitration law not being my Mastermind specialist subject, could use some reassurance on this regard. Cheers - Nosebagbear (talk) 23:42, 22 February 2023 (UTC)
I understand the idea, but despite the explanations, I am not sure of the implications of this new clause and share the concerns. In the event of claims by a group of users in relation to the damage caused, does the Foundation have the power to force an unwilling UPE to arbitration? (I'm not a lawyer, but I'm afraid not). Conversely, in the case we apply the policy "Don't think, and erase all", could the UPE come and complain, arguing that we made it lose money by following this policy? Or by revealing the names of their clients? I'm afraid so. And if their lawyer's fees are very high, am I wrong thinking they know that they can be reimbursed if the mediator decides that such an article or such a sentence should be reintroduced ? Or that the name of their clients should not be revealed, for example in the name of "business secrecy" ? Additionally, could Wikipedians fighting UPEs be considered one of the parties? Just asking because I know I could not afford the fees splitting. --Pa2chant.bis (talk) 07:10, 23 February 2023 (UTC)
Thanks for the excellent reply. I was falsely assuming that UPEs would be suing us. Looks like it's the other way around. Sounds like it boils down to "WMF sometimes wants to sue UPEs, but in the past it has been cost-prohibitive to do so. If we force them to do mediation, and to pay for the legal costs, then this makes WMF taking legal action against UPEs a new, practical possibility." Sounds good to me! 👍👍 –Novem Linguae (talk) 22:47, 22 February 2023 (UTC)

I don't quite understand what the aim of mediation is. If it turns out that someone has violated our terms of use, we block the account and discuss if the contributions should be deleted. What is the point of mediation? What would be a situation where mediation would make sense and what would be the result? If WMF reaches an agreement with a company in such a mediation, would the community have to accept contributions from that company? --Holder (talk) 08:15, 23 February 2023 (UTC)

I've tried to address this question in response to a similar question above under the heading: "1. The way UPE currently works is paid editors ignore the ToS, astute Wikipedians spot them and create a sockpuppet investigation." No, we would never be forced to accept contributions from anyone. I've addressed this now below in greater detail under the heading "1. response to questions about whether or not this would affect editors ability to change content". Thanks for the feedback. SSpalding (WMF) (talk) 18:09, 23 February 2023 (UTC)

Probably a rather stupid question, but am I right to assume that the parties involved in such a mediation will be in any case WMF on the one hand and a UPE company on the other hand? And, to repeat Holder’s question, what will be the aim intended by WMF? Some sort of cease-and-desist declaration from the UPE company which in case of future violations makes it easier for WMF to sue them? --Jossi2 (talk) 16:58, 23 February 2023 (UTC)

In addition, many UPE spammers reside in third world countries where the legal system shows varying degrees of dysfunction. How will the mediation result be enforced? Will this new system be effective in these circumstances? MER-C 17:33, 23 February 2023 (UTC)
Good question. See my response above under the heading "3. One interpretation is that the WMF seems to think it can compel anyone it suspects of UPE to travel to California..." There's no guarantee that we would be dealing with a jurisdiction with predictable and functional legal systems. This would be a problem with litigation in the same way it's a problem with mediation. There's actually a huge benefit to mediation in dysfunctional legal systems because we don't have to rely much on the formal legal system to make it work.
Said differently, litigation requires us to work within a country's unreliable legal system through the entire complaint, judgement, and settlement enforcement process. Mediation would only require us, in theory, to rely on that country's legal system for the "settlement enforcement" process. This is another one of those "cost saving" and "time saving" things I allude to in my response above under the heading: "2. If so, why would mediation be superior to other processes?". I hope that addresses your question! SSpalding (WMF) (talk) 17:59, 23 February 2023 (UTC)
It does help and is better than the current process - although I am not convinced that it will be effective against third world spam. MER-C 20:17, 24 February 2023 (UTC)
Noted. We will take that into consideration to assist us with for the final wording as well as in our day-to-day choices for how to enforce the language (should it be supported by the community) SSpalding (WMF) (talk) 18:58, 28 February 2023 (UTC)
No worries. There are no "stupid questions" and many of your questions overlap with @Pa2chant.bis @Indy beetle and @Nosebagbear's questions so I will try to address everyone here before my next meeting :p . I think something that might address all of the content related questions above:
1. response to questions about whether or not this would affect editors ability to change content
No, it would not. The mediation would determine a very specific question: did this company breach the contract (TOU) by what they did, yes or no?
If yes, then we can legally compel them to stop violating the rules and hold them responsible through a court order.
If no, then the company can continue using the projects, but the community and/or the Foundation's ability to stop their actions doesn't change if either believes they have violated a rule again. It also doesn't change the disposition of the community on the previous rule violations.
Either way, article content wouldn't be locked or frozen. A mediator couldn't reverse a community edit/takedown. There would not be a hold or limitations on editing if the company "won" the mediation. Old community rule enforcement edits would not be reverted. Finally, a marketing company successfully defending against our accusation could never end in a situation that would grant a company the power to violate the TOU or community rules in the future. What happens to article content after it's done would never be at issue.
To give a simplified analogy: imagine a company gets audited by the IRS for suspected tax evasion for the year 2010. During the hearing, the IRS doesn't find enough information to prove the company violated tax rules. That does not give then the company free reign to violate tax laws from 2010 forward now that they've "won" against the IRS. It also doesn't prevent the IRS from auditing the same company again in 2015 if another rule violation is suspected.
Finally, the IRS would never be in a position where a judge could force the IRS to change the tax laws only for that specific company moving forward because the company survived the audit. This is an inelegant analogy, but hopefully it makes it more clear explain why a pay-for-play situation where a company gets preferential content treatment during a settlement is not the type of thing we could be forced by the mediator to do.
2. what will be the aim intended by WMF?
Any settlement with a marketing company would likely include:
A. Stop all of the activities that are against the rules that you're currently engaged in.
B. Promise you won't do the same or similar things in the future under penalty of a court order.
C. Disclose the comprehensive list of (i) all accounts you've used and (ii) all companies you've worked with (so we can provide misinformation back to the community to do any necessary future clean up and investigation)
Assuming after mediation may renege on their promises, then we can take the settlement to be enforced in a court in the company's jurisdiction. A firm specializing in selling UPE services would see this new language as an existential threat.
3. Additionally, could Wikipedians fighting UPEs be considered one of the parties?
No, the TOU is a contract between the company that agreed to it and the Foundation. In mediation, there is no "impleading" (where someone could be "brought in as a party"). There's also no "calling witnesses" etc. Those are two more reasons that make mediation more viable than litigation in these kinds of disputes. SSpalding (WMF) (talk) 17:50, 23 February 2023 (UTC)
Thank you for clearing that up SSpalding. -Indy beetle (talk) 18:05, 23 February 2023 (UTC)
Thanks for your answers @SSpalding (WMF), appreciated.
Presumably we (or, more likely, either SPI-equivalents or CUs, would need to provide the original evidence - which WMF CUs may or may not duplicate for confirmation later), with that evidence wanting to demonstrate that x editor had broken the rules and there was reasonable reason to think that x editor worked for y company?
It wouldn't need to be in the TOU, but assuming it goes in (and I can't see any reason it shouldn't), it would be useful for Legal to be able to say at some point what they are looking for...and possibly what would be counterproductive. I mention the latter, for example, because there are things like "mystery shopping" that I believe fr-wiki uses, amongst others, and would be useful to know if that would render a request for Legal to take this action in a specific instance more or less likely. Nosebagbear (talk) 18:44, 23 February 2023 (UTC)
Noted. Interesting that you bring up fr-wiki because there's no shortage of targets for exactly these types of mediation actions that they've already identified on Antipub The type of information on a page like that may be the catalyst to identify the marketing companies involved in the activity who could be targeted. The Foundation also regularly gets messages notifying us of meaningful problems related to UPE especially when it's a scam (someone stole money from someone who didn't know better that Wikipedia pages weren't for sale). If I understand your comment correctly, I would confirm that we'd likely give general statement (possibly in a Diff post that people can refer to) about what information might be helpful in these types of actions. Hopefully, that would guide people to reach out to us if they see a problem that they believe is connected to a wider network of undisclosed paid editors. SSpalding (WMF) (talk) 18:56, 23 February 2023 (UTC)
Dumb question: can volunteers - or the WMF on our behalf - seek compensation from undisclosed paid editors for cleaning up after their edits via this means? MER-C 20:17, 24 February 2023 (UTC)
Thanks SSpalding (WMF), that sounds great.

Some questions:

  1. Is there precedent for a binding med-arb in website TOS as opposed to just arb?
  2. Same question about requiring only a half-day or full-day: is that a reasonable limitation?
  3. Same question about requiring teleconferencing... is there legal precedence for this? What if the party doesn't have the technology required for a teleconference?
  4. Is this the first time a major website has put a mandatory one-day teleconference med-arb clause in its TOS or has this been done before? (It's a cool idea, but seems novel.)
  5. The second paragraph seems to be one-sided... will the WMF also agree to timely provide any documentation, etc.?
  6. Why doesn't the agreement specify who will run the med-arbitration, how many arbitrators, what rules will govern the proceedings, etc.?
  7. Is any of this enforceable against non-US persons? Should there be one clause for domestic parties and another for international, to increase its scope/enforceability?

Thanks, Levivich (talk) 00:25, 2 March 2023 (UTC)

Also, a sidenote, but I couldn't figure out if the TOS was w:browsewrap or w:clickwrap, I guess the latter is intended, but FWIW I noticed that when I edit using w:WP:WIKIPLUS, it doesn't have a notice saying by publishing I agree to the TOS, and I don't see one on the account registration screen at enwiki. Thought I'd mention that. Levivich (talk) 00:32, 2 March 2023 (UTC)
  1. Many people philosophically do not like arbitration. Also, arbitration is more expensive and has more procedural requirements than mediation. For both reasons, we thought the best approach would be to default to mediation (with a negotiated settlement) before the matter turned into a arbitration (which would force settlement) on any remaining open issues. Med-Arb is unique to online platform TOUs, since other platforms don't have ethical barriers around forcing their users into arbitrations or cost barriers against litigation. That said, there is lots of precedent for mediations escalating toward arbitration in U.S. contract law.
  2. This is both reasonable and standard. U.S. mediations are almost always half day (4 hour) or full day (8 hours). This stipulation is mostly to avoid the other side suggesting that the mediation be stretched out inappropriately (e.g. two weeks of one hour sessions) to artificially raise our costs or otherwise subvert the process.
  3. Generally, commerical platforms force companies in disputes with them into going physically to a specific jurisdiction. For various reasons, we wanted to avoid this and allow any dispute to be handled to be decided in the location of the company's choice (i.e. via teleconference). We believe that this lowers a barrier making it more likely that the company will voluntarily comply rather than needing to use extra judicial process to force the company into a specific jurisdiction through court orders.
  4. Teleconference is common when alternative dispute resolution is suggested. "Arbitration" is very common. Med-Arb is uncommon. So some of this is novel, but then again, a lot of the way Wikimedia projects work doesn't (and shouldn't) take cues from other commerical platforms. 🙃
  6. There is an implication to act in "good faith" in all U.S. contract law. So if we ask someone to mediate under the contract, we have to at least act a minimally to make sure that the company isn't unduly prejudiced. Should we not do that, then the other side could have their own recourse against us. Even if this weren't the case, the mediator is a neutral third party. They would require a mediation brief to be filed with them which, usually, is accessible (in some form) to both parties. It would be a very bad look to have intentionally withheld information in that brief when the point of the process is supposed to be a good faith settlement. (To be continued)
SSpalding (WMF) (talk) 01:44, 2 March 2023 (UTC)

I have one unaddressed concern about "Marketing Company Mediation": the matter of getting the other party to actually respond to binding mediation. I admit I'm not a lawyer, but from my experience with online communities it can be difficult if not impossible to match an individual with a username, especially if that person makes a determined effort to obfuscate the connection. And even if the party responsible is identified, it will often be a lone individual -- either working on his or her own, or as a straw man hired as a contractor -- who has no incentive to participate in binding mediation: said person will simply respond to any legal action with "That's not me, & you have to prove it is. And anyway, what are you going to do to make me respond?" They may not even be in the US, let alone a country that offers remedies to make them respond. Further, while at the beginning perhaps only a few will succeed in evading identification -- & thus cannot be served to respond -- as the strategies on their part are refined, it will be increasingly difficult to identify these people: this has been the case with spammers, trolls & other malcontents on the Internet. I suspect the Foundation may find itself resorting to litigation in every case to either force them under penalties to stop their editing, or to extract money for damages & punitive penalties. -- Llywrch (talk) 08:37, 7 March 2023 (UTC)

I also have a question, SSpalding (WMF). If the other party can be court ordered to show up to mediation, and they do show up, and they say "Hey, screw you, we're going to keep doing what we're doing and we're not changing a thing", is there anything that anyone can do about that short of actual litigation? Seraphimblade (talk) 06:31, 17 March 2023 (UTC)

Good question. Iy has a straightforward answer. With a court order, the only legal question in front of the judge would be, "does the terms of use apply to X company or not?" Since it would be really unlikely for us to pursue any claim where it wasn't completely obvious that the terms of use applied to the Company, I imagine we would win many more of these than we'd lose.
Now to answer your question, the Company could reject that. But in countries where there is strong rule of law, Companies typically don't ignore court orders. It's really unlikely, for example, in the U.S. or Singapore that a business legally operating in that jurisdiction would simply say "no," to a court because it opens up ongoing penalties that might even be able to be enforced by law enforcement in certain scenarios. For any reputable business in a jurisdiction with strong rule of law, the decision between either violating a court order and showing up to a video conference mediation like they've been instructed to would be a no brainer.
That said, there are more likely ways one of these companies could challenge a court order like this like appeals and delay tactics. The thing that might discourage this is the fact that the losing party has to pay attorney's fees. So if the other side inappropriately delays and adds cost, they will know there's a possibility they might have to pay both their costs and our costs related to the delay by the end of the process. SSpalding (WMF) (talk) 16:44, 17 March 2023 (UTC)

Mandatory arbitration is often of dubious enforceability. There are countless cases in various jurisdictions. What jurisdictions has the Wikimedia Foundation considered, and how do we know this is going to reduce costs and increase certainty rather than the opposite? This proposed clause seems designed for enforcement against businesses in the USA and poorly thought out for individuals and anyone outside the USA. Nemo 10:30, 31 March 2023 (UTC)

Jurisdictions that WMF has directly considered for future actions have included the U.S., one specific E.U. country, and Ukraine. The link you referenced describes, correctly, that the general rule is that arbitration requirements are not "null and void" in the EU even in the face of insufficient drafting. The article cites a handful of lower court cases (that didn't reach the appellate level) as interesting specifically because they don't seem to follow that general rule. In other words, this nullification is an anomaly that contradicts, "long and settled line of authority on the arbitrability of mandatory principles of EU law," as the article describes.
That said, your point is definitely taken that every jurisdiction will present its own difficulties as to whether or not this new language will be useful. There are a handful of other discussions during the earlier days of this consultation about what our tactics would be working in a jurisdiction where we'd unlikely be successful. The short answer is that we almost certainly wouldn't move forward in those jurisdictions because there are already currently a handful of "good" cases in jurisdictions where we'd be much more obviously successful. In conclusion, this measure doesn't claim to fundamentally fix an intractable problem like UPE, but we think it helps in many ways that manner. SSpalding (WMF) (talk) 17:40, 3 April 2023 (UTC)

Hello there,

I am still incredulous as to the impossibility of initiating legally binding proceedings against marketing accounts. First of all, the foundation does have enourmous financial resources, as well as other means of influence (implied by its enormous viewership), to hold perpetrators to account. Then, I strongly suspect that in most countries of interest, the guilty party would have to pay any legal fees. Finally, I don't see how a legally binding process such as arbitration would not be preferred.

At this moment, it seems to me that the foundation is on the brink of giving up on holding illicit marketing accounts accountable, and takes away a resoundingly popular method of at the very least inducing legal worries in the perpetrators. I believe that WP is full of illicit marketing accounts, and that what the foundation needs are better, stronger and less subvertible tools of ensuring that all editors are legitimate. --Mathmensch (talk) 11:09, 22 April 2023 (UTC)

Broad WMF flexibility of interpretation

Regarding the lines We reserve the right to exercise our enforcement discretion with respect to the above terms. Where required, enforcement of these terms may include actions not listed in the Wikimedia Foundation Office Action Policy. If enforcement is required in new circumstances, we will make an effort within at most one year to update the Office Action Policy to catalog the new type of action, I have a mixed bag of questions and comments

  1. "enforcement of these terms" appears to specifically mean "we can take additional unspecified actions to act against a user, but not additional unspecified grounds on which to act" - is that interpretation correct? It goes off the same definitional split that the UCOC and UCOC enforcement guidelines are using.
  2. "We will make an effort" seems rather vague - I'm aware that with a need to create an operational viewpoint and get legal signoff timelines can run, and perhaps longer still if needing to seek functionary and/or community advice, but it feels like within a year we should either have an addition or have had it declined (at least, declined for future actions "we did it once, it wasn't ideal, not again")
  3. The reasoning [side note - big fan of the whole about section] states This is to account for new requirements under changing laws. - is that the WMF saying that it doesn't take actions outside the Office Action Policy (or won't) except as required by changing law? It won't carry out actions before adding them to the OAP in order to resolve issues generated from other reasons?

Cheers, Nosebagbear (talk) 17:28, 22 February 2023 (UTC)

@RamzyM (WMF): - it looks like different staff members are responsible for answering different aspects of the TOU, so I wasn't sure who to ping on the UCOC/conduct sides. Would it be possible to get answers for the above? Nosebagbear (talk) 12:08, 10 March 2023 (UTC)
Sure, please let us get back to you next week. RamzyM (WMF) (talk) 13:05, 12 March 2023 (UTC)
This one was flagged to me for reply. I think these are good questions. Let me go through the list and I'm curious for thoughts after I do so if you still think we can improve this language. I think questions one and three go together: the idea being that the Office Action Policy is supposed to describe what actions the Foundation directly takes and is an effort to provide transparency about what we do so that people aren't surprised. I can think of a few reasons we might take actions outside of the Office Actions Policy:
  • One, as noted in the third question, is if legal requirements (new or existing) result in us being forced to do something even though we didn't describe it in the office actions policy. This is the one we're trying to anticipate with this wording. We want to be clear to people that if we have the legal requirement to act, we will need to do so even if the Office Action Policy has become outdated, but if that happens we'll work to get the Office Action Policy updated within a year to account for the new requirements.
  • Two, we might act outside the Policy if the communities asked us to do so. For example, if there is a new topic area that becomes similar to child protection work and functionaries as a group ask the Foundation to take on new work, I think we could do so at the same time as we worked on an update to the Office Action Policy to reflect the change, rather than having to wait until the policy update is done to change how the work gets done.
  • Three, if we identify some sort of new and very severe harm that requires emergency action, we might act followed by a policy update afterwards. I can't think of any realistic examples of this though. An unrealistic one might be if someone started using some sort of Wikipedia data to target a series of bomb threats and the Foundation found out and deleted several things to help stop the emergency. The point of this section is that if we had to take such an emergency action, we are committed to updating our policies within a reasonable time to reflect that need in the future. Or if it was a mistake, to realizing the mistake and working with the communities to have a better way to respond to such situations to avoid the mistake repeating.
Lastly, the second question about the language. We drafted it that way to account for varying legal processes that may take longer than a year. It's also good legal drafting practice for a Terms of Use document to be more general and avoid a firm commitment that might not make sense in some unknown future scenario. It's the hardest document to change of all the Foundation policies, so hard committing to a timeline regardless of future events rather than saying that we'll do our best to stick to the timeline would honestly not be the best legal work for the Foundation. That said, I personally agree that we can and should make at least some kind of temporary conclusion within a one-year timeframe if this kind of situation comes up, even if the final conclusion might require an appeal in a court case that ends up taking several years. I'm open to suggestions on how we can word the commitment here to meet the dual goals of being legally flexible and ensuring good transparency about Foundation actions. -Jrogers (WMF) (talk) 23:03, 12 March 2023 (UTC)

"volunteer editors who must", and the link behind it

I see what this is supposed to mean, but it reads like an oxymoron. Also, linking to a non-authoritative information page, a non-policy, a non-guideline, from the Terms of Use is not ideal. Just link to en:WP:COI instead... if a link to a specific wiki's page from the global TOU is needed at all. ToBeFree (talk) 18:15, 22 February 2023 (UTC)

Just saw this suggestion and I like it. We'll likely use this link for our second draft, coming soon. -Jrogers (WMF) (talk) 23:07, 12 March 2023 (UTC)

Accessibility edits

A small point, but it would be really good to change several links to follow accessibility guidelines.

Instead of "you can contact us [link]here[/link]", it is much better to expand the link text to be more like "you can [link]contact us here[/link]". This means the eye is drawn to the meaning of the link rather than just the word "here"; similarly, assistive technologies can (for example) read out a list of links rather than a list of "here, here, here" and so on.

Thanks! — OwenBlacker (Talk) 19:20, 22 February 2023 (UTC)

Thanks for this suggestion! -Jrogers (WMF) (talk) 18:12, 23 February 2023 (UTC)

Applicable legislation

The phrase I want to discuss is:

Please be aware that you are legally responsible for all of your contributions, edits, and re-use of Wikimedia content under the laws of the United States of America and other applicable laws (which may include the laws where you or the subject you’re writing about are located live or where you view or edit content).

While I understand the new phrasing is more inclusive, as a non-native English speker, I also consider it less clear. I suggest the following alternative:

Please be aware that you are legally responsible for all of your contributions, edits, and re-use of Wikimedia content under the laws of the United States of America and other applicable laws (which may include the laws where you live, where you view or edit content, or where the subject you’re writing about is located).

Strainu (talk) 19:22, 22 February 2023 (UTC)

"Where the subject you're writing about is located" is incorrect, though. Russian law is not applicable to an article about Putin, for example. --Rockstone35 (talk) 00:12, 26 February 2023 (UTC)

capitalization

Attorneys love to capitalize words that feel important, but there are reasons why this should be rejected in favor of downcase. First, there's no grammatical reason to capitalize what's not a proper noun, or the first word of a sentence, in English. Second, and more importantly, capital letters are not pronounced. When they are used as a form of emphasis or to hint at importance of a term, that inference is lost on 1) anyone who cannot see, 2) anyone who otherwise is reviewing the terms in an audio format, and 3) anyone who understands how English capitalization works and doesn't understand that attorneys sometimes use capital letters for reasons unrelated to grammar and orthography.True Pagan Warrior (talk) 19:45, 22 February 2023 (UTC)

Useful observation. Some things that are capitalized are purely discretionary: For example, "Empower and Engage" is in title case. For those, I've made a note to think carefully about capitalization choices. Other things that are capitalized are not as discretionary. For example, "APIs" is a "defined term," so anytime we need to refer to it we have to capitalize it as a technical matter. Finally, something like "Privacy Policy" is both a defined term and referring to the specific project Privacy Policy (rather than privacy policies in general), so for both reasons, it's capitalized. If you have specific words that would make things more clear, then very open to suggestions. If not, your point is still well-taken and we'll consider this when revising. SSpalding (WMF) (talk) 21:33, 22 February 2023 (UTC)

Creative Commons imports and sui-generis database rights

First, let me say I consider the move to CCBYSA 4 as a huge step forward for content reuse! Happy to see this change which will open troves of data for reuse.

However, section 7, points a seems to go beyond the license requirements. It says says that for the text the user owns, it's possible to reuse the facts from the projects without attribution. As far as I understand (from section 4), the CCBYSA 4 license does not allow such reuse without the attribution.

This raises 2 issues:

  • a moral one: we're weakening the copyleft part of the license
  • a technical one: with the current attribution rules of most projects, it's virtually impossible to correctly identify imported vs original facts 100% of the time, which makes such reuse impractical on the large scale (read: with bots)

I suspect this phrase was introduced to help with the porting of data from Wikipedia to Wikidata, but it seems to leave us in the same spot as before. Perhaps we should consider other options that do not weaken the SA clause. Strainu (talk) 20:26, 22 February 2023 (UTC)

See #Sui generis database rights above. I agree with your concern about attribution. The licence should require reusers –
  • to attribute facts gleaned from Wikipedia to Wikipedia at least in some minimal way, and
  • to share alike, to retain the viral aspect of the CC licences rather than having the information locked away in proprietary boxes.
Andreas JN466 00:43, 23 February 2023 (UTC)
To attribute facts? That's a bit silly. If I bring something up in conversation and I happened to learn it from Wikipedia, do I have to include a notice of attribution? What if I'm not sure where I read it? Also, facts per se are not covered by copyright, unless I greatly misunderstand. No one should own a fact or rights over someone's knowledge of that fact on the basis that the person learned it from a certain source - we only have rights to our actual writing, not all understanding which can from it be gleaned. And Wikipedia is supposed to get its facts from sources anyway. CharredShorthand (talk) 03:12, 23 February 2023 (UTC)
Attributing individual facts is not required by the license either. What is required is attributing when reproducing "all or a substantial portion of the contents of the database". For example, when bulk-moving information such as the population for settlements from a whole country from Wikipedia to Wikidata.
Don't get me wrong, I support such information reuse, but in many countries (all of Eastern Europe, at least) attribution is still considered critical for reusing open gov data. The EU just recently started encouraging countries to use CC licenses for releasing such data, before it was even worse. Strainu (talk) 08:00, 23 February 2023 (UTC)
Exactly. Perhaps it'll be helpful to reproduce here WMF Legal's assessment from some years ago (m:Wikilegal/Database Rights):
The sui generis right is infringed when "all or a substantial part of the database" is transferred into another document or database. Unfortunately, the Directive does not define substantial, but does say that "substantial" is "evaluated qualitatively and/or quantitatively". [1] Extracting and using a insubstantial portion does not infringe, but the Directive also prohibits the "repeated and systematic extraction" of "insubstantial parts of the contents of the database".[2]
When there is large-scale extraction of Wikimedia data the volunteer community should be credited somewhere. --Andreas JN466 08:38, 23 February 2023 (UTC)
Indeed, facts are not copyrightable, so Andreas' demand that The licence should require reusers [...] to attribute facts gleaned from Wikipedia to Wikipedia at least in some minimal way is rather pointless, since the license is about our copyrights and can thus only require (i.e. withhold permission for) things that we have a copyright for. Regards, HaeB (talk) 08:20, 25 February 2023 (UTC)
You're talking past us. It is abundantly clear from what has been said above that this is not about individual facts, but about "repeated and systematic extraction". Andreas JN466 16:24, 25 February 2023 (UTC)
I was specifically talking about your own suggestion for "the licence" (I think you meant ToU) where you used the word "glean" (Wiktionary: "To gather information in small amounts, with implied difficulty, bit by bit"), and I was merely adding ("Indeed") to the concerns that others had already raised about your this suggestion of yours. See the #Sui generis database rights section above for my more general questions about your apparent interpretations of the database rights issue. Regards, HaeB (talk) 17:52, 25 February 2023 (UTC)
No, I did indeed mean the Wikidata licence. On English Wikipedia, for example, the WMF added a little notice some time ago to the Edit window, saying: "You agree that a hyperlink or URL is sufficient attribution under the Creative Commons license." This is in practice a slight modification of the CC BY-SA 3.0 licence, to make attribution less cumbersome for re-users. What I had in mind was that a similar note could be added to a CC BY-SA-licensed Wikidata as well, to specify a non-cumbersome method of attribution.
By "gleaning" I was simply referring to the activity of bots importing Wikipedia infobox contents, category systems, interwiki links and so forth. To glean = "to gather (something, such as information) bit by bit, to pick over in search of relevant material", per Merriam Webster. Andreas JN466 00:27, 26 February 2023 (UTC)
Sorry, now you are suddenly talking about a CC BY-SA-licensed Wikidata, i.e. you propose to relicense all of Wikidata from CC0 to a CC BY-SA license? That wasn't on the agenda here and it's also not how one could have reasonably understood your original proposal above (The licence should require reusers ...).
But even setting that aside and assuming that you were instead talking about a project like Wikipedia whose (text) content is indeed under a CC BY-SA license, there seems to be a disconnect (probably also related to your fundamental misunderstanding about the difference between the 3.0 and 4.0 licenses regarding database rights). Sure, you could conceivably be right that adding such a little notice to the edit window amounted to making the content available under a license that is slightly more permissive than the CC BY-SA 3.0 license that is "officially" stated in the site footer (by allowing reuse with a form of attribution that would not be sufficient under the original license, if that is indeed the case). But we can't use that mechanism to make the license more restrictive by adding requirements such as your two bullet points. Regards, HaeB (talk) 09:28, 26 February 2023 (UTC)
That's what happens when you carry on a conversation spanning two sections. You asked me above what I'd want to be done about Wikidata. Given the "Share Alike" aspect of Wikipedia's licence, the Wikipedia and Wikidata licences would come to the same thing if the special waiver is rejected and Wikidata systematically extracts data from Wikipedia, as it surely has been doing.
So: A Wikipedia published under CC BY-SA 4.0 would require Wikidata to be published under the same licence, requiring attribution and Share Alike compliance. And the type of attribution required for Wikidata could be specified in a similar way to how the attribution burden was lightened through that edit notice for Wikipedia, so it does not become too cumbersome. It could perhaps be something as simple as displaying the phrase "Powered by Wikidata" to the end user, combined with a meaningful link to Wikidata. Don't you agree it would be worthwhile having our re-users acknowledge the vast community effort ("sweat of the brow" indeed ...) that went into these projects? Andreas JN466 12:14, 27 February 2023 (UTC)
1. You continue to be fundamentally mistaken about the differences between 3.0 and 4.0, see my detailed explanation above.
2. I don't recall asking you would want to be done about Wikidata, that's a rather poor excuse for hijacking the conversation about this particular ToU change to advocate for your grand project of relicensing all of Wikidata over a decade after its launch. ...
3. ... Rather, I had asked you twice before to explain what you mean exactly by phrases like Wikidata systematically extracts data from Wikipedia or large-scale, systematic imports of data from Wikipedia. You have consistently failed to do so, not even providing just one concrete example. This makes it difficult to discuss how much these extracts might really be affected by legal restrictions (i.e. whether Wikidata has really been violating them for many years), and what the benefits and drawbacks of the proposed ToU update might be in that regard.
4. Your suggestion that attribution could perhaps be something as simple as displaying the phrase "Powered by Wikidata" is in stark contrast to your own claims above that lack of attribution for data(base) contributors to our projects like a government body or other organisation would be an end run around the actual intent of the 4.0 license, and a bullying move. So rather than a simple "Powered by Wikidata" label, under your own logic we would have to imagine that as such a reuser, one might have to display a list of possibly tens of thousands of past data donors to Wikidata, extracted from billions of Wikidata edits, so as to not "bully" some provincial European town who a decade ago uploaded a list of their council members or such.
Regards, HaeB (talk) 17:48, 28 February 2023 (UTC)
1. You are not listening. CC BY-SA 4.0 requires attribution in cases that were not (or weren't clearly) in the scope of CC BY-SA 3.0. You say those cases are meaningless because 3.0 did not licence re-use in those cases anyway. I say it is not meaningless, because 4.0 provides a new tool to demand attribution and sharing-alike in those cases which were previously uncovered by the CC BY-SA 3.0 licence. Having such a new tool seems valuable to me, given that there is a lack of readily available alternative means to demand attribution and sharing alike from re-users.
2. The questions you asked me above were, "Do you think that Wikidata has been (or is) in violation of database rights for Wikipedia, or not? If yes, what precisely is the "substantial part" being "harvested"? Also in that case, would you really prefer having to shut down Wikidata for legal reasons over going through with the proposed ToU update?"
3. I didn't answer because the question struck me as absurd. When I wrote that Whither Wikidata? piece in The Signpost, more than half of all citations in Wikidata were to Wikipedia (and phrased in the most generic way, i.e. "Italian Wikipedia", "English Wikipedia", without linking a specific article version of even a specific article, for that matter). These were all Wikipedia imports.
4. I said "Powered by Wikidata" combined with a meaningful link to Wikidata, just like the Google Knowledge Graph panel just displays the word "Wikipedia" with a relevant link. I am not interested in listing data contributors. I am interested in having us – the WMF and the community – find ways to enable users to track provenance, because I believe it's vitally important for a healthy and socially beneficial information ecosystem.
I think at this point I don't feel like discussing this topic with you any further. It seems to me we have exhausted the potential for any possible benefit to be derived from this discussion by either of us. This is not to say that I haven't derived any such benefit, just that the discussion has run its course. I thank you for your comments and hope you know that I value and appreciate your deep knowledge of Wikimedia affairs and their history. Regards, Andreas JN466 21:12, 28 February 2023 (UTC)
For reference and clarity, the full text of this change in section 7 is:
Where you own Sui Generis Database Rights covered by CC BY-SA 4.0, you waive these rights. As an example, this means facts you contribute to the projects may be reused freely without attribution.
So first, the example sentence is, while technically accurate, a bit misleading IMHO: As discussed above, facts are not copyrightable and can thus already be reused freely without attribution except when they are covered by restrictions other than copyright, here database rights. These only come into play in special situations and certain countries, but not e.g. for almost all non-bot edits on Wikipedia, where this distinction is thus immaterial. It does seem that everyone in this thread is aware of this. But general readers of the ToU might well come away with the mistaken impression that facts would generally need to be attributed except for this waiving.
Anyway, wording issues about the example aside, the legal team gave a rationale for the first sentence at m:Terms_of_use/Creative_Commons_4.0/Legal_note#Waiving_database_rights and a blog post series linked there, which expands on the difficulties they want to avoid with this. (And no, porting of data from Wikipedia to Wikidata is not mentioned there as far as I can see.)
Regards, HaeB (talk) 09:28, 26 February 2023 (UTC)
The legal note says: The proposed amendment to the Terms of Use waives potential database rights (where applicable) covered by CC BY-SA 4.0 to keep barriers low when sharing information.
That's simply not achieved, not as long as reusers need to dig through all the history of an article in order to determine whether the content was imported (therefore the exemptions does not apply) or was otherwise created before or after the new ToS applies. This also raises the risk that the data will be reused in good faith without attribution and some rights holders will decide to enforce the license.
With all the dual licensing and such Wikipedia reuse is already a mess, why make it an even bigger one? Strainu (talk) 11:09, 27 February 2023 (UTC)
  1. ↑ See Database Directive art. 7(1), (2)(a). The term "quantitatively" "refers to the volume of data extracted from the database and/or reutilised, and must be assessed in relation to the volume of the contents of the whole of that database". The term "qualitatively" "refers to the scale of the investment in the obtaining, verification or presentation of the contents". Case C-203/02, Brit. Horseracing Bd. Ltd. & Others v. William Hill Org. Ltd, 2002 E.C.R. I-10415, §§ 70 et seq.
  2. ↑ Database Directive art. 7(5).

triggering of epilepsy: seriously?

"Posting or modifying content with the intention to seriously harm others, such as deliberate inducements to self-harm, or deliberate triggering of epilepsy." Unless it's some US norm saying that this text must be included in any ToU, this sentence sounds science fiction to me. I have very strong doubts that any text written on any Wikimedia projects would push someone to hurt himself. Any person able to self-control will not do self-harm (and the ones that cannot control themselves probably are on the wrong website).

Same for triggering epilepsy. It is written that some combination of images and sound (e.g. in videogames) have more chances to trigger epilepsy, but nothing coming from any media will trigger a crisis for sure. Thus this sentence has no sense.

I thus suggest to point to a different direction. What we want to avoid is to willingly posting of modifying content with the intention to seriously harm others, such as deliberately spread false information, or hateful contents. Manipulating information is central when we talk about hurting people, and protecting minorities. What is meaningless is try to prevent actions that have no scientific basis or that assume a very high level of stupidity from users. Ruthven (msg) 21:34, 22 February 2023 (UTC)

  • The epilepsy thing is a plausible hypothetical, but I agree it's an odd one. I would say the old English Wikipedia essay "Beans" applies to that one. "Deliberate inducements to self-harm" is an actual phenomenon that happens on nastier corners of the internet and in other settings, so it seems an applicable example. I don't think language specifically against "deliberately spread false information, or hateful contents" would cover anything not already in the UCOC, plus what constitutes "false" or purely "hateful" information is often hashed out in volunteer discussions, so dragging the WMF more directly into that would give them a greater role over content I don't think anyone is asking for. -136.56.145.246 22:27, 22 February 2023 (UTC)
    This aspect reads as specifically handled to handle directly injurious content - regarding inclusion, a reason is given in the about section. In terms of whether it actually happens, epilepsy charities deal with posted media attempts to do it all the time on their social media, and I believe an attempt was made in the last two years on en-wiki's epilepsy page. Guaranteed, not at all, even for the 3% of those with photoepilepsy, but wilful attempts are made.
    I'm also confused @Ruthven: by but nothing coming from any media will trigger a crisis for sure. Thus this sentence has no sense. - someone who posted a flashing media at a common photoepilepsy frequency on en:Photosensitive epilepsy without a reasonable legitimate reason has the intention to deliberately trigger a seizure - whether they succeed or not doesn't undermine their intention. So it remains coherent and retains its sense. Nosebagbear (talk) 23:37, 22 February 2023 (UTC)
    Just to add, re photoepilepsy triggering - those of you in the UK may have been following the Online Safety Bill debates, including this recent development: https://epilepsysociety.org.uk/what-we-do/campaigns/zachslaw . WP:BEANS is an important point to consider, though. We'll do so. PBradley-WMF (talk) 12:40, 23 February 2023 (UTC)
    Ok, Ok, if triggering epilepsy is plausible, then lets forbid it. But I feel that this is a matter that shouldn't be covered by the terms of use: posting a flashing media in order to trigger photoepilepsy would be out of the scope of the project anyways, and treated as a regular vandalism by our admins. In the same way, I reckon that inciting self-harm on Wikimedia pages is also completely absurd, and would be considered something close to a personal attack by the admins. The case of Conrad Roy, who was incited to kill himself by his girlfriend, is very specific, as its emotional context. Nothing to do with usual interactions among users (hopefully). Discussing suicide (or self-harm) on Wikimedia projects is just using the wrong platform. This is why I think that focusing on such specific aspects (that are already covered by other rules, the main one is "being in scope", which is already covered by 1. Our Services, 3. Content We Host and 4. Refraining from Certain Activities) "distracts" from what we really want to forbid. Namely: Uncivil interactions between users (already covered by previous points), and using the platform for attacking minorities (or any group in particular, it shouldn't be a minority finally). This is why I suggested to rephrase the sentence to aim at these aspects. Ruthven (msg) 13:19, 23 February 2023 (UTC)
    Sure, and the grounds of "it's already prohibited" is far more reasonable and I'd agree with it, if it hadn't been for the reasoning on the about page. Although other sections, too, would seem to cover the areas you note. Nosebagbear (talk) 15:26, 23 February 2023 (UTC)
  • I also thought it was weird, and also thought of beans. Bluerasberry (talk) 00:42, 23 February 2023 (UTC)

It might be helpful to give the respective explanation in the "About"-page a read on this matter: "[...] we added a line regarding contributing content that’s intentionally designed to hurt a reader such as by causing them to harm themselves or triggering an epileptic seizure. This was already prohibited under the more general terms, but we believe it is likely that upcoming laws, the UK Online Harms Act in particular, will require an explicit reference to this. DBarthel (WMF) (talk) 13:47, 23 February 2023 (UTC)

So what does this mean for Commons? I think we may have a couple of these flashing images somewhere. Do we need to delete them or can we keep them if they were not uploaded with the intention to trigger anyone (but to satisfy the curiosity of non-affected people)? --El Grafo (talk) 08:10, 2 March 2023 (UTC)

The discussion has unfortunately derailed into the epilepsy distraction. But the real threat here to encyclopaedic work is «Posting or modifying content with the intention to seriously harm others». This term enables the Foundation to basically delete any contribution and kick out any user for almost any reason, as long the ban gets couched in «oh, we thought the edit could harm».
Hypothetical: we are living under Lysenkoism, and a user makes an edit supported by WP:RS that challenges the Lysenkoist scientific consensus; Wikimedia Foundation can happily argue that such an edit can cause serious harm to others because it is anti-science, anti-agriculture, and so the edit is harmfully pro-mass starvation.
Lysenkoism was the scientific consensus in the Soviet Union for 35 years and 3,000 geneticists who happened to disagree were shot to death.
Geocentrism was the scientific consensus of his day and poor sap Galileo Galilei was hounded into "admit[ting] that, contrary to his true intention, a reader of his Dialogue could well have obtained the impression that it was intended to be a defence of Copernicanism" Galieo was harming the reader! Galileo spent the rest of his life under house arrest. Me, I don't wanna get banned from Wikipedia for disagreeing with the "scientific consensus," much less shot or house arrested for life. XavierItzm (talk) 21:10, 5 March 2023 (UTC)
The concept of "posting with the intention to seriously harm" is so ambiguous as to be an inappropriate condition for participating in Wikipedia.
Take as an example an article discussing of things that people do that affect climate change. Those who believe in climate change would say that removing text about limiting one's carbon footprint would have "intention to seriously harm". There exist millions of people who have the (likely mistaken) belief that humanity will be extinct in a hundred years unless we do such things, and extinction is certainly a harm. However, those who think that climate change is an unproven theory can make an identical claim that posting text suggesting that people limit their own travel, avoid commuting long distances, turn down their thermostats, etc, is causing harm to anyone who follows it. (Their life has fewer options, they have more illnesses, they are limited to less fulfilling or less lucrative jobs, etc; and all for no actual societal benefit; this is a serious harm.) I have had an edit reverted for merely suggesting in a climate-change page that climate models "predict" what "might" happen rather than "calculate" what "will" happen in the future. Many people who treat climate-change as a religion or belief think that anyone who casts doubt upon it is by definition harming the planet.
Similar issues can arise on any topic that has strong partisan sides. For example, articles that inform Ukranians about methods used to repel Russian forces may have "intention to cause serious harm" to Russians involved in the war, while article text that informs Ukranians how some have peacefully surrendered to Russian attacks may have "intention to cause serious harm" to Ukranians. Ditto the Middle East wars, the abortion controversies, religious beliefs, etc.
Even in a topic very close to the probable motivation for this phrase (suicide), there are factions with strong and contrary beliefs. Out on the global web, articles that describe reliable ways to kill oneself have been posted by the en:Church of Euthanasia, which believes that the planet is overpopulated. These were not posted with the intention to seriously harm anyone, yet they were censored by legal threats after someone actually followed them successfully. Similar medical texts describe how the unnecessary pain and suffering of terminal patients can be relieved by assisted suicide. However, people who believe that life is sacred would claim that making such information available causes harm and is posted with the intent to cause harm.
The inclusion of the word "intention" makes it even more problematic. The same identical text in an article could be perfectly legitimate if the intention is to help (eg a pregnant woman), but could be censored if the intention is to harm (eg an innocent fetus). Who decides what the intention of the poster was? Someone who perhaps has a completely different and polarized PoV from the poster? This is an unworkable term in a ToU for a project with as wide a scope as Wikipedia. Gnuish (talk) 21:07, 13 March 2023 (UTC)

Oxford comma, or lack thereof

Under Management of Websites, the new ToU lists "if such correspondence was made in bad faith, repetitive, unfounded, and/or abusive" but in most other places (such as "do not create an employment, agency, partnership, joint control or joint venture relationship" under Other terms) there is no Oxford comma. This is truly a trivial point but I do believe formatting should be be consistent within the document. Thanks, Novo Tape (talk) 21:44, 22 February 2023 (UTC)

  • In retrospect, I should probably add a list of examples on the revised ToU, so I will. Under the summary,it says "violate copyright, post illegal content or violate other laws" which is without an Oxford Comma. Ditto for "You adhere to the below Terms of Use, the Universal Code of Conduct and to the applicable community policies." Under overview, there is " contribute, monitor, or delete content" which does use an oxford comma."a contributor, editor, or author" It also says "posting, modifying or reusing" "medical, legal, or financial issues" "growth, development and distribution" "offensive, erroneous, misleading, mislabeled, or otherwise objectionable" "medical, legal, or financial issues" "threats, stalking, spamming,vandalism or harassment" "harassment, exploitation, or violation" "Knowingly accessing, tampering with, or using" "a statement on your user page,a statement on the talk page accompanying any paid contributions, or a statement in the edit summary accompanying any paid contributions" "page footer, page history, and discussion page" "availability, accuracy, or the related content, products, or services" "Detect, prevent, or otherwise address" "Refuse, revert, disable, or restrict access" "bad faith, repetitive, unfounded, and/or abusive" "be safe, secure, uninterrupted, timely, accurate, or error-free "directors, officers, employees, and agents" "direct, indirect, incidental, special, consequential or exemplary" "damages for loss of profits, goodwill, use, data, or other intangible losses" I believe these are all the instances where an Oxford comma should or should not have been used, depending on the guidelines. I retract my earlier statement and instead believe that an Oxford comma should be added to those few places where there is not one already. Novo Tape (talk) 21:33, 23 February 2023 (UTC)

Automated Editing

With the rapidly increasing complexity of OpenAI type software, I think that not including a clause about disclosing the usage of such a program is a mistake. I also feel a clear distinction should be made between "using bots" and "using an OpenAI".

DarklitShadow (talk) 00:19, 23 February 2023 (UTC)

I don't see the need for using legal tools. Each Wikiproject should be able to make their own rules about what disclosure they require for AI usage. ChristianKl ❪✉❫ 20:28, 23 February 2023 (UTC)
I agree with ChristianKl. I think this is best handled by local policies. Nosferattus (talk) 00:33, 3 March 2023 (UTC)

Limitation on Liability

This is legal language capping lawsuit limitations at $1000.

Something about this is not synced with other WMF legal programs, offers to help, and offers to intervene.

I can understand lawsuit limitations for readers being capped, but where the WMF, Wikipedia, and Wikipedia editors intersect, the editors have asked for and expected more WMF liability. Sometimes wiki editors have been harassed, exposed to violence, arrested, detained, or had other problems as a result of typical Wikipedia use. The Wikimedia Foundation has the Legal Fees Assistance Program to help with some of this - not the harassment part for example - and although the Wikimedia Foundation might be framing this as a possible favor to editors the Wikimedia community would like this as a commitment and liability from the WMF as an organization.

Going further - I think it is the case that sometimes WMF programs encourage Wikimedia community members to take risks and put themselves in harm's way by organizing Wikipedia programs and making themselves more visible than they might have done otherwise without WMF staff intervening in community activities. WMF's diversity programs especially, while well intentioned, do not have third party ethics oversight and may have different rates of attracting harassment than Wikimedia community organized outreach programs for the same demographics.

I wonder what the distinction is between the terms of use of Wikipedia as a website and WMF staff interventions with users of the website. With sites like Facebook and YouTube, their terms of use apply to a community in which almost none of the users interact with company staff. In Wikipedia a relatively high percentage of content contributors interact with WMF staff in personal ways. I routinely question the safety of WMF staff encouragement for users to do things that their peers would not recommend doing, like running outreach programs with higher risk of negative attention.

I am not a lawyer but I think the principle here is en:moral hazard. WMF is claiming a limitation on liability, while staff personally go into the platform and encourage users to increase their risk, and there is no community advocate reviewing the WMF interventions. When WMF staff encourage community to increase their risk then I wish there were another commitment somewhere that WMF increases their liability for such cases. Bluerasberry (talk) 01:04, 23 February 2023 (UTC)

Thank you for flagging this concern Bluerasberry . I think in this case, you do actually have a misunderstanding of the relevant law here. The Terms of Use are the legal contract for people using the websites. A limitation of liability like this section protects the Foundation from legal claims arising from the sites as covered in these terms and makes the community editing model safer. Absent this kind of waiver, a reader or article subject could argue along the classic reliance on a windmill crankshaft case that they relied on something on Wikipedia under the ToU and then it changed or was deleted and it caused them harm. Nearly all contracts have a limitation of liability provision as a result of that case to make clear that some unknown reliance by others won't cause huge damages. If we didn't have a section like this, the risk is that the Foundation would have unexpected risk urgent situations where we'd have to intervene on project content to limit liability. On the other hand, the examples you give (legal fees assistance and staff-coordinated activity) aren't affected by this provision at all. They are separate legal issues not covered under the contract represented by the ToU. -Jrogers (WMF) (talk) 18:25, 23 February 2023 (UTC)
@Bluerasberry: where the WMF, Wikipedia, and Wikipedia editors intersect, the editors have asked for and expected more WMF liability - I don't recall seeing such demands, do you have links documenting such a community consensus about the need to be able to sue WMF for sums larger than $1000? Regards, HaeB (talk) 09:36, 24 February 2023 (UTC)

Mission

"We offer websites and technical infrastructure to help you do this."

The Wikimedia Foundation's current social infrastructure is vast and highly influential. I wish that social infrastructure were invested in the Wikimedia community Global Council and I interpret this statement to be thoughtful in limiting the Wikimedia Foundation to technical infrastructure as it moves away from social infrastructure.

Comments on why social infrastrucure is not mentioned here? Specifically, is this thoughtfully and intentionally phrased to focus on technical infrastructure? Bluerasberry (talk) 01:09, 23 February 2023 (UTC)

The intent behind this is simple: the 2012 language mentioned "websites" but not APIs. The APIs are now increasingly important ways that knowledge on the projects is transmitted. As a result, the "technical infrastructure" addition is a nod to the other mentions of the API later in the document. It is definitely not intended as a comment on any of the social structures the Foundation supports or the Global Council. SSpalding (WMF) (talk) 01:27, 23 February 2023 (UTC)

Refraining from Certain Activities, removal of reference to information being false

Seems strange and counterproductive to remove "that is false or inaccurate" from the subsection "Engaging in False Statements, Impersonation, or Fraud" - especially since that was the only use of the word "false" in the very section addressing false statements. Mahonhughes (talk) 01:20, 23 February 2023 (UTC)

Indeed that proposed change raises many questions:

With the intent to deceive, p Posting or modifying content with the intent to deceive or mislead others that is false or inaccurate

Notably, so far this rule of the Terms of Use (TOU) has set out an objectively verifiable criterion, namely the posting of false or inaccurate content (the actus reus), linked to a further criterion related to intent as a mitigating factor in the sense that good-faith-edits should be excused. The proposed amendment, on the contrary, attempts to define the offending activity by the underlying intent. Whether a posting is, or is not, against the TOU would thus depend solely on the intent. Thus the (suspected) intent itself, rather than an actus reus, would be the criterion for determining whether the TOU have been overstepped. Under such a provision, anyone could easily be accused of acting against the TOU (alleged intent), and at the same time such a provision remains a toothless tiger when it is not possible to demonstrate the underlying intent. In summary: The TOU should hinge on verifiable criteria related to the edit itself, not just on intent. See below for a suggestion as to how this could be formulated. --Chris Howard (talk) 13:13, 18 March 2023 (UTC)

With the intent to deceive or mislead others, posting or modifying content such as to result in content that is false or inaccurate

This is spread out over too many documents

In particular the many newly inserted references to a "Universal Code of Conduct" would require someone to navigate back and forth far to many times to figure out what the rules actually are. Get rid of UCoC and just say what the rules are all in one place. --ManaUser (talk) 07:10, 23 February 2023 (UTC)

I do agree that UCoC should be direct attached to document or ask the reader to read UCoC before read this document Meganinja202 (talk) 07:52, 23 February 2023 (UTC)
Worth noting if you were taking this position (I don't render a viewpoint either way), you'd also need to include things like the office action policy and the privacy policy within it Nosebagbear (talk) 15:30, 23 February 2023 (UTC)

Licensing and attribution of Wikisource translations

In 2018 a very substantial translation of mine, bearing my real name, was deleted from Wikisource, without my being contacted, because it bore a GFDL license rather than a CC BY-SA 3.0 or 4.0 license.

If 3.0 licenses are discontinued in favor of 4.0 + GFDL, will provision be made to automatically upgrade current 3.0 pages to 4.0 + GFDL, so that more pages are not deleted without the authors' knowledge; or at least to notify the authors of planned deletions?

Is there a way I could consult appropriate staff about any possibility of restoring my deleted page? When I contacted a Wikisource staffer, I was informed that its resuscitation would require substituting "Translated by Wikisource" in place of my real name. I'm willing to contribute Wikipedia texts anonymously, but I feel that translators should, if they wish to, be allowed to contribute translations to Wikisource under their real names, and I would be glad to present my arguments for that.

Could I please be contacted at my Wikipedia talk page, if there should be any responses to my questions?

Thank you.

Nihil novi (talk) 07:42, 23 February 2023 (UTC)

Some Questions and feedback of mine

Questions 1- In overview, what would be a "urgent threats of serious harm" ? Also, I find this part very vague and TOO MUCH open to interpretation, i think it should be more objective and clear, to avoid unfair decisions

Feedback: 1- In overview, "we generally require that all content" should be kept

2- in 1a, users like yourself should be keept or edited so the friendly part of text is kept

3- In 1b, keep WMF I think that it is clearer this way

4- In 10,

We reserve the right to suspend (temporarily, or permanently)

Please, Remove permanently or make the language more clear about the theme, WMF should never leave the door open to turn off the possibility to future FAIR questions or responses to reports, as it is right now, I fear it would open door to unfair and authoritarian decisions in the future

If anything be free to talk with me in my Wiklipedia Page or replying this, also have a nice day Meganinja202 (talk) 08:01, 23 February 2023 (UTC)

I agree with all of this. Carpimaps (talk) 10:13, 23 February 2023 (UTC)
Thanks for the feedback here, we can look back through some of the small changes. The last point about being able to refuse to receive reports is actually in the Digital Services Act. It's something it allows any web host to do whether we put it in the ToU or not, primarily to protect against repeat spammers and bad faith requesters. The Foundation does, unfortunately, receive emails like that, where someone tries to lie to us repeatedly to change something they don't like. The purpose of the language in the ToU is to inform people about how it works so they know not to send spam or fraudulent demands. -Jrogers (WMF) (talk) 17:58, 24 February 2023 (UTC)
I agree with all of this Kalesh (talk) 15:20, 28 February 2023 (UTC)

Full support for move to CC-BY-SA-4.0

I offer full support for the move from Creative Commons CC-BY-SA-3.0 to CC-BY-SA-4.0. For some background on my thinking, please see:

  • Morrison, Robbie (February 2023). "Which open data license?". openmod forum. Open Energy Modelling Initiative. Retrieved 2021-05-06.  Evolving blog.

In particular, version 4.0 of that license provides much better legal provisions for data and datasets than version 3.0.

The remainder of the changes all look very sensible too. Good luck. RobbieIanMorrison (talk) 08:36, 23 February 2023 (UTC)

@RobbieIanMorrison Thanks, interesting. What do you make of the waiver that the WMF would like to add to its Terms of Use when 4.0 is adopted? (The proposed wording is "Where you own Sui Generis Database Rights covered by CC BY-SA 4.0, you waive these rights. As an example, this means facts you contribute to the projects may be reused freely without attribution.") Would it impact provenance tracking? Would it impact the ability to import data from EU databases? Regards, Andreas JN466 14:02, 27 February 2023 (UTC)
The waiver makes sense in my opinion. Requiring attribution for data is a pernicious idea and creates a slew of problems. Provenance tracking can be handled by other (non-legally-binding and more flexible) means. We're already restricted from importing data from non-freely-licensed EU databases, which are the vast majority, so the effect of this change will be small (and won't affect Wikidata since it already sensibly requires CC0). Nosferattus (talk) 15:25, 3 March 2023 (UTC)
I am not going to enter the debate about CC0‑1.0 versus CC‑BY‑4.0 for numerical data relative to provenance, veracity, overhead, and practicality. This matter also depends on the use cases in question and the wishes of the data provider. But the Sui Generis wording highlighted by Jayen466 looks very odd. 96/9/EC database rights only project against substantial extraction under certain circumstances, namely substantial investment in infrastructure and genuinely impaired commercial risk — and the the CC‑BY‑SA‑4.0 naturally waives that protection while adding the requirement for attribution (the "BY" trait). Another problems is determining whether Sui Generis rights actually apply — the presence of copyright is often clear, but not so for 96/9/EC rights. I suggest that the WCF revisits that highlighted phrase. RobbieIanMorrison (talk) 09:35, 14 March 2023 (UTC)
I strongly oppose the use of CC-BY-SA-4.0 or any of the other 4.0 licences for the reasons already expressed in m:Talk:Terms of use/Creative Commons 4.0#General support and opposition back in 2016. The 4.0 licences exist to undermine the rights of authors at a time when the original author's "ownership" of their work is already being undermined by duplicate content penalties. The 4.0 licences contain an ugly loophole which, instead of taking the ability to use the content away from anyone infringing on the terms of the licence, once and for all, instead gives them a thirty-day grace period to say "oh gee, I forgot" and continue business-as-usual without being forced to take the content down. That's a problem because even getting a word with an intellectual property lawyer starts at $300/hr or more. Your opponent wastes a hundred hours of your senior counsel's time with excuses as to why they should be allowed to steal your content, that'll be $30000 please... and most of the garden-variety local lawyers insist that, while they do offer business law, this isn't business law but intellectual property law - some sort of specialised legalistic black art that requires a specialist big-city lawyer to even address. Individual wiki authors can't afford this, and the big for-profit commercial wiki farms who want to steal your content know this. If CC 2.0 or 3.0 required the infringing content be taken down, while 4.0 cynically keeps giving second chances, that's a licence to infringe and steal content. Then there's the "personality rights" issue and the sui generis database question. Once again, 4.0 exists to take rights away from authors. This is the sort of thing which could happen:
  • Someone creates a wiki for travel. Clever idea if it makes the data any easier to update when it's outdated and wrong, but travel is a desirable segment for advertisers and two founding members of the project sell the domain for an undisclosed sum (which turns out to be in the $1.4 million range) and the new owner runs the project into the ground. The users, understandably, are dissatisfied so they take their content and go elsewhere. The various CC-SA licences expressly permit this, but instead of complying, the for-profit company launches frivolous lawsuits against the original contributors. I'd presume (and this is hypothetical and not legal advice) that CC 2.0 would allow the takedown of the for-profit site for violating the share-alike provision, while 4.0 would give the alleged infringer an undeserved second chance.
  • Someone creates an article about Boy Scouts in a wiki encyclopaedia; a local scoutmaster makes the well-intentioned decision to upload a photo of his/her local scout troop. Someone else, posting at a for-profit wiki farm, decides this image would be great to illustrate an article on a wiki eroticising spanking. The problem should be obvious; maybe the photographer is too stupid to realise that slapping any CC licence on anything is pretty much taken as an open invitation to steal the content and ignore the licence conditions (and the licence doesn't prohibit this conduct in any case) but none of the people appearing in that photo consented to that self-serving commercial use. A group of online trolls threatening to screenshot the whole mess and send it to the commercial wiki farm's advertisers might be able to get the content taken down, but the actual photographic subjects? They get no say, and CC 4.0 makes the problem worse by taking away personality rights. Never mind that the author (the photographer) has no right to waive those rights as they belong to the subject (the people appearing in the photos), it'll be $300+ for every hour of counsel's time to address this when the infringers hide behind this ugly "free" licence. The average boy or girl scout doesn't have that sort of money, even if the motto is "be prepared".
  • Someone creates a clever wiki, maybe anything from an "un" encyclopedia as a parody to a fan's collection of star trekking memorability. They host this outside Wikipedia under one of the non-commercial licenses like CC-NC or CC-BY-SA-NC 2 or their variants. Again, whomever owns the domain name goes behind the backs of the community and sells out to a for-profit commercial wiki farm, who runs the project compliantly initially, but gets greedy after a half-dozen years and slips some really nasty text into their site's terms and conditions contracting the wiki farm out of the non-commercial terms of the licence - the big print says "CC-NC" but the fine print slips under the radar. Edit anything and you've just reposted it under an incompatible licence. Very deceptive, but there's a problem... the user who edited that page today isn't the original creator and has no right to waive the CC-NC conditions. Again, this is a big company screwing over a bunch of little, unrepresented authors. Attempt to enforce your rights, expect to be ruined by insane legal fees. After paying a few hundred dollars an hour (and it could take $30000-100000 to bring this to trial, money the victims don't have) do you expect the content to be permanently taken down (as CC-NC 2.0 would have done) or the infringing party merely slapped across the wrist and given thirty days to correct the issue (the CC 4.0 model)?
These aren't hypotheticals. To make matters worse, if the original author attempts to exercise their right to take their content and go elsewhere, the search engine duplicate content penalties will ensure that no one will ever find anything but the infringing version if it is on what was the original domain name. That destroys the right to fork, in all but name.
It's almost as if these people were going out of their way to harm authors by taking away their rights. I have no idea why. Are they being lobbied by the commercial wiki farm sites who think that people like me were put on this earth to be their unpaid employees? Are they merely opposed to all copyright in principle, and see CC as a forum to take this crusade out on the little guys while the big media companies continue "all rights reserved" as business-as-usual, where anything from out-of-print materials to the work of dead authors remains tied up in copyright for decades to come?
I do not consent to my work being relicenced under any of the 4.0 licences. If you do proceed with this, please remove anything I've contributed over the years and any derivative work. Thank you. Carlb (talk) 22:14, 14 April 2023 (UTC)

Minor copyedit

In section 4, the first list reads as:

Engaging in harassment, threats, stalking, spamming, or vandalism or harassment as described in the UCoC;

Transmitting chain mail, junk mail, or spam to other users.

Posting or modifying content with the intention to seriously harm others, such as deliberate inducements to self-harm, or deliberate triggering of epilepsy.

The second full stop should be changed to another semicolon Grimmchild (talk) 09:22, 23 February 2023 (UTC)

vollkommen unverständlicher Passus / completely incomprehensible passage

Dieser Passus ist vollkommen neu in den Benutzerrichtlinien:

> Unbefugtes Hochladen von technischen Ressourcen Dritter
> * Hochladen von technischen Ressourcen Dritter unter Verstoß gegen die „Richtlinie zu Ressourcen Dritter“, die von den technischen Teams der Foundation verwaltet wird.

Der Autor/Übersetzer verkennt hier vollkommen, was für einen Otto-Normalwikipedianer verstehbar ist. Letzterer hat vielleicht mit Mühe gelernt, dass man Text bei Wikipedia "Bearbeiten" kann. Schließlich klickt er auf den so bezeichneten Button. Ob es jedem klar ist, wenn das als "Hochladen" bezeichnet wird, kann bezweifelt werden. Manche wissen zudem, dass Hochladen auch für Bilder, Audios und Videos geht. Aber warum erklärt der Autor nicht für jeden verstehbar, was er unter technischen Ressourcen versteht und wie man die bei Wikipedia hochladen kann? Nur wenn man weiß und versteht, was nicht erlaubt ist, kann man es vermeiden. Offenbar war der Autor des Abschnittes noch nie auf einem Stammtisch mit Normalwikipedianern oder ist Coder und kann sich nicht in jemand hinein versetzen, der das nicht ist. Tirkon (talk) 11:39, 23 February 2023 (UTC)

This passage is all new in the Terms of Use:

:> Unauthorized uploading of third-party technical resources. ::> Uploading third party technical resources in violation of the "Third Party Resources Policy" managed by the Foundation technical teams.

The author/translator completely misses the point here, what is comprehensible to the Average Joe Wikipedian. The latter may have learned with some effort that one can "edit" text on Wikipedia. After all, he clicks the so designated button. Whether it is clear to everyone that this is called "uploading" can be doubted. Some also know that uploading also works for images, audios and videos. But why doesn't the author explain in a way that everyone can understand what he means by technical resources and how to upload them to Wikipedia? Only if you know and understand what is not allowed, you can avoid it. Obviously the author of the section has never been to a meeting with regular wikipedians or is a coder and can't put himself in the shoes of someone who isn't. ) Tirkon (talk) 11:39, 23 February 2023 (UTC), translated by DBarthel (WMF) (talk) 11:33, 24 February 2023 (UTC)

Otto-Normalwikipedianer liest sich ja auch üblicherweise die Nutzungsbedingungen durch und wundert sich darüber, dass er Begriffe darin nicht versteht. ToBeFree (talk) 17:45, 23 February 2023 (UTC)
The Average Joe Wikipedian usually reads through the terms of use and is surprised that he does not understand terms in it too. ToBeFree (talk) 17:45, 23 February 2023 (UTC), translated by DBarthel (WMF) (talk) 11:33, 24 February 2023 (UTC)
I'm sorry this section was confusing. We had noted concern about it from some comments above as well and are currently looking at how we can rephrase it to be more clear. -Jrogers (WMF) (talk) 18:02, 24 February 2023 (UTC)
Es tut mir leid, wenn dieser Abschnitt verwirrend war. Wir haben auch in ein paar Kommentaren oben Bedenken dazu gefunden und prüfen gerade, wie wir ihn umformulieren können, um ihn klarer zu machen. Jrogers (WMF) (talk) 18:02, 24 February 2023 (UTC), translated by DBarthel (WMF) (talk) 11:13, 27 February 2023 (UTC) @Tirkon, @ToBeFree

The cancel of the alternative disclosure policy page

In a nutshell: Removing the need for a decision to turn a local paid contribution policy into an alternative policy explicitly and specifying it on a dedicated page, may change the status of those that have not been defined in this way until now.

Hello, I would like to point out a problematic change.
In the proposed change, it is suggested that the following condition be removed from the terms of use:

An alternative paid contribution policy will only supersede these requirements if it is approved by the relevant Project community and listed in the alternative disclosure policy page.

Currently, (at least on the Hebrew Wikipedia) there is local paid contribution policy. The policy was not defined as an alternative paid contribution policy, and were not listed on the alternative disclosure policy page. It is only an additional local policy that the community has defined like any other local policy, so it has never been determined that it will replace the policy of Wikimedia Foundation from the terms of use.
Removing the quoted sentence as currently proposed may retroactively change the status of such policies, prevent them from adding to the terms of use without detracting anything, and causing them to override the terms of use from now on, without the community determining that it is interested in this. —מקף⁻ණ‎‏ (Hyphen) 14:36, 23 February 2023 (UTC)

حذف کلمات انگلیسی، عربی و بیگانه

در ویکی پدیا پارسی کلماتی بیگانه وجود دارد که معادل برای مثال شرح=توضیح، منبع=سرچشمه و غیره. حتی کلمه‌ی فارسی درست نیست و پارسی درست است که شما میتوانید با یک جست‌وجوی ساده به آن پی ببرید. من درخواست دارم تمام این‌گونه کلمات حذف شوند و معادل پارسی آن جایگزین شود و به بنده کمک شود. Amz86 (talk) 14:54, 24 February 2023 (UTC)

گمان می‌کنم این درخواست شما و به صورت کلی استفاده یا عدم استفاده کلمات عربی یا انگلیسی در فارسی بشدت بحث‌برانگیز است.
I think this request of yours and encouraging or discouraging usage of Arabic or English words in Persian is very controversial. Farooqkz (talk) 19:06, 4 April 2023 (UTC)
پارسی‌سازی را خودمان می‌توانیم اندک‌اندک انجام دهیم، ولی مشکل بزرگتر این است که ویکی‌پدیای پارسی بیش‌ازاندازه به ویکی‌پدیای انگلیسی وابسته است. دیری نمی‌گذرد که ویکی‌پارسی به 1 میلیون جستار برسد، ولی بخش بسیار گسترده‌ای از الگوها و پودمان‌ها همواره بر مبنای ویکی‌پدیای انگلیسی هستند. نمی‌دانم این‌همه شیفتگی به ویکی‌پدیای انگلیسی برای چیست، ولی بهتر است کمی مستقل‌تر باشیم.Darknessswamp8 (talk) 20:47, 7 April 2023 (UTC)

You must make that disclosure in at least one of the following ways:

  • a statement on your user page,
  • a statement on the talk page accompanying any paid contributions, or
  • a statement in the edit summary accompanying any paid contributions.

One of the concerns I had when this was implemented was that someone checking whether a given edit was paid needs to check all three locations because of the "at least" requirement. Could the WMF elaborate on whether this is still fit for purpose, and if so explain why? MER-C 17:29, 24 February 2023 (UTC)

Interesting. Would you say that requiring disclosure in all three places would meaningfully relieve burden on editors checking for disclosures? SSpalding (WMF) (talk) 22:20, 24 February 2023 (UTC)
Not just editors, but also the public. This is important information for determining the trustworthiness of Wikipedia content. In addition, there have been anecdotes of paid editors removing disclosures from their user pages months after the edits were made. MER-C 11:31, 25 February 2023 (UTC)
Thanks. That's really useful information. SSpalding (WMF) (talk) 16:33, 25 February 2023 (UTC)
I would give a big +1 to this. HouseBlaster (talk) 20:53, 25 February 2023 (UTC)
Agree with MER-C as well. Andreas JN466 14:33, 28 February 2023 (UTC)
Totally agree with MER-C. Perhaps we could also add that every account used for paid editing should be of the form User:Example (PAID). In general, I don't see paid editors complying with the requirements as stated, and those that do "disclose" do so in a very confusing or incomplete way. Smallbones (talk) 19:49, 28 February 2023 (UTC)
Agree with Smallbones as well, thanks to him.
Interesting. Community local policies would need to be updated should this be enforceable in the context of on-wiki contributions.
That said, in theory if rules like that were in place, it does create a bright line that would give editors and readers much less work to review contributions for proper disclosure.
When we present the revised text, I've made a note to myself to see if there's any way to deal with "incomplete or confusing" disclosures within the limited scope of the TOU. SSpalding (WMF) (talk) 20:36, 28 February 2023 (UTC)
The easiest way to deal with this would be the following text"

You must make that disclosure in at least one of the following ways:

  • a statement on your user page,
  • a statement on the talk page accompanying any paid contributions, or and
  • a statement in the edit summary accompanying any paid contributions.

The individual projects should then be able to introduce the User:Example (PAID) as an equivalent acceptable way of more conveniently doing the same thing. It would also get rid of the problem of undeclared (and almost undeclareable) paid IP editing. Smallbones (talk) 03:28, 2 March 2023 (UTC)

An article talk page disclosure would still be required, even with a special user name format, if disclosure in all three locations were mandatory. isaacl (talk) 17:43, 11 March 2023 (UTC)

Is there any evidence that this part of the Terms of use has improved the quality of participation in the wikis? Nemo 06:59, 28 February 2023 (UTC)

I have found frequent occasion to refer to its language in dealing with paid editors, and would find it much simpler to spot snd explain if all 3 were required./ DGG (talk) 10:04, 2 March 2023 (UTC)

Also agree with this. The problem for a reader is that they need to check three different locations - page history, page talk and the user page of each contributor - to determine if an article has involved paid editing. The only place they are likely to check is the page talk, if that, so mostly they won't realise. From an editor's perspective, knowning that a particular edit was paid for is helpful, especially if you are otherwise unaware that the editor is paid, because if it wasn't mentioned there you would need to check the userpage and the article talk page to find out, and thus it is likely to be missed. From an admin's perspective, having a list of pages that were edited for pay on the user page gives a very useful overview - otherwise, once again, you need to check every article talk page they have edited or look at their list of contributions and try to find it in there. Having all three covers all three use cases. - Bilby (talk) 22:36, 2 March 2023 (UTC)

If we're going to require disclosure on the User page and the Talk page, requiring it in the edit summary as well seems excessive, IMO. If we're going to stick with the "at least one of the following" version, however, I think it makes sense. Nosferattus (talk) 00:30, 3 March 2023 (UTC)

Suggestion/Question: The European Union and various states and institutions have created "Lobbyist Registries" (or Register for Transparency), usually with a form to fill out. Maybe we could take inspiration from it? => Would it be smart?, useful?, practical? to have paid editors complete a declaration form? …with some mandatory minimum information (identity of contributor, Client, Objective, date or duration of assignments, Financial resources available), and a free “other information” box).
Ideally the person should also undertake not to publish under IP.
Once this form has been validated by the registrant, a mention could automatically appear on the user's page and on his talk page, as well as on each explanation of modification in the history of the articles to which this person contributes.
Related Question: This sounds a lot more complicated and tricky to me, but should we also consider the case of people who voluntarily perform "commissioned work" on behalf of entities with a commercial and/or political purpose?.--Lamiot (talk) 21:43, 6 March 2023 (UTC)

Are we not all voluntarily writing in some extent for political reasons? Because writing for 'free access to information' is still a valid political statement.
Doppeltracktion (talk) 10:04, 8 March 2023 (UTC)
Question By requiring that the user page be one of the places where paid editing must be flagged, are you implicitly assuming that there will not be questionable edits coming from anon-IP addresses, such as edits to The Kluger Agency (a product-placement firm) from every business hotel in the land or edits to j2 Global (a fax-to-email company) from anon-IP address blocks which are owned by that company? These sort of patterns are only going to get harder to spot if WMF drifts away from MediaWiki's pattern of identifying unregistered users by displaying their bare IPv4 addresses, which seems likely at this point. Carlb (talk) 22:59, 14 April 2023 (UTC)
No, it would mean that in order to comply with the paid-contribution disclosure requirement, you need to have an account. No assumption is being made about who makes questionable edits. isaacl (talk) 05:28, 15 April 2023 (UTC)

Licensing change in overview

In the Overview section, there’s a change:

[…] we generally require that all content you contribute is made available under a free license or in released in the public domain.

This is

  • confusing. [C]ontent you contribute is made available under a free license or in released in the public domain – which one? How do I know whether content I contribute will be made available under a free license or it will be released in the public domain?
  • I think legally problematic. Now you don’t require that content is available under a free license, but rather you make it available under a free license, whether its copyright owner did so previously or not. What happens legally if they didn’t?

Tacsipacsi (talk) 02:12, 25 February 2023 (UTC)

Minor consistency and formatting notes

  • The summary includes the full email info@wikimedia.org, section 10 contains info[at]wikimedia.org, and section 8 info@wikimedia.org.
  • Section ten states Detect, prevent, or otherwise address fraud, false or unverifiable information security, or technical issues [...]. I think a comma is missing and the passage is intended to read Detect, prevent, or otherwise address fraud, false or unverifiable information, security, or technical issues [...].
  • Section 7(c) is set to read in part in such case you warrant that the text is available under terms that are compatible with CC BY-SA (or, as explained above, another license when exceptionally required by the Project edition or feature). This sounds ungrammatical to me. Maybe something more like this is intended: in such cases you warrant that the text is available under terms that are compatible with CC BY-SA license (or, as explained above, another license when exceptionally required by the Project edition or feature).
  • The terms editor and author are removed from the Our Terms of Use, but are set to still appear in the Overview section (join as a contributor, editor, or author, a team of contributors, editors or authors. Sections 1, 4, 7 are also set to still use the term editor as well. Sections 7 is set to still use the term author

—The Editor's Apprentice (talk) 08:25, 25 February 2023 (UTC)

Accidently submitted early, here are two more notes:
  • The phrase licensed or qualified is removed from the overview, but still appears in section 3.
  • The phrase live or where you view or edit content is removed from the overview, but still appears in sections 1 and 4.
—The Editor's Apprentice (talk) 08:29, 25 February 2023 (UTC)

Commons and paid editing

  1. I see that "An alternative paid contribution policy will only supersede these requirements if it is approved by the relevant Project community and listed in the alternative disclosure policy page" has been struck, but "A Wikimedia Project community may adopt an alternative paid contribution disclosure policy. If a Project adopts an alternative disclosure policy, you may comply with that policy instead of the requirements in this section when contributing to that Project" remains. I can't make sense of how that affects Commons, which does not currently require disclosure of paid editing. What is the intent here, and what problem is being solved by striking that sentence?
  2. The new text, "In addition, if you make a public posting on a third-party service advertising editing services on Wikipedia in exchange for pay, you must disclose what Wikipedia accounts you will use for this service in the public posting on the third-party service": is this intended to be specific to Wikipedia (as against other WMF projects) or was that just sloppy wording?
Jmabel (talk) 20:41, 27 February 2023 (UTC)
1. Commons is one of the largest projects with a distinct policy that does not require disclosure. We want to honor that. If a marketing company is following the rules of the project, we want to make it clear that they're not violating the TOU by doing so. Said differently, imagine that there are (1) on-wiki project rules, and (2) the terms of use. A bad actor may violate one or both. With that remaining language we want to say that should a company act in a way that seems like a violation of #2 (TOU), but it turns out to acceptable under #1 (local community rules), then we won't consider it as a violation of the TOU.
2. In drafting, we have tried to narrow the scope as much as possible before presenting this to the community. We want to make sure that the community is on board with the minimal amount of changes to achieve the maximal effect. The vast majority of systematic UPE problems relate to Wikipedia (especially because, as you noted, Commons has less strict policies). So we limited it to the more clear and present issue. If you support broadening the language to (for example, "Wikimedia projects") then that'd be a good point to know. SSpalding (WMF) (talk) 18:18, 28 February 2023 (UTC)
@SSpalding (WMF): it's mostly that from a Commons point of view, these changes have been very confusing. People there thought the striking of "An alternative paid contribution policy will only supersede these requirements if it is approved by the relevant Project community and listed in the alternative disclosure policy page" meant that Commons no longer could have such an alternative policy. Glad to see that they were wrong. - 07:18, 1 March 2023 (UTC)

Avoiding violate

Can we get rid of the term violate, please? To my mind, it supports a strongly legislative moralism point of view. As if law was always legitimate, and not blindly following law was necessarily an infamous act made through bodily or mental violence, which is always comparable to rape in term of gravity and ignominy.

Skimming through the already present comments, I see that how some references to totalitarian states, so I guess that by now arguments were already well exposed to express a divergent theory that, yes, law can be the illegitimate tool of perfidious oppression.

I see the term infringe was already mentioned in discussions too, but I think contravene would be a better option here. Psychoslave (talk) 22:44, 27 February 2023 (UTC)

I subscribe to the proposition to replace violate with contravene. 2dk (talk) 20:12, 9 March 2023 (UTC)

Click “help” on the left side of most pages

Actually, we don't know if the user is gonna click a mouse, touch the smartphone screen or press entry on its keyboard after due keyboard focus.

Let's put apart that in left-to-right languages, your link might just as well be on the right side, translators might think about it after all. Still, we don't know what skin the user prefers, or if the user is actually using a screen reader for which case spatial orientation is meaningless.

As an alternative Follow the link labelled "help" which is present in the main menu of most pages could do the trick. Psychoslave (talk) 22:54, 27 February 2023 (UTC)

+1 to this. Additionally the "Help" link is hidden in a menu by default on Vector 2022, and not in the menus on the mobile website at all. the wub "?!" 16:10, 4 March 2023 (UTC)
+1. In addition to what you said, having to click help “on most pages” is a lot of work. --80.153.205.25 08:39, 21 April 2023 (UTC)

API Terms, "..which are incorporated into these Terms of Use by reference"

The proposed TOU incorporates 3 wiki pages by reference. Will any editing restrictions be placed on those pages as a result? Obviously someone shouldn't be able to vandalize the page and that automatically be part of the TOU, but I think requiring any change to those pages require going to a burdensome review process (vaguely defined) would be pretty annoying. Legoktm (talk) 05:06, 28 February 2023 (UTC)

Relatedly, there's supposedly a project to improve API policy documentation, is that supposed to be done before the TOU update and supersede the 3 listed policies? Mostly asking because the linked Robot policy especially needs some work, so it would be nice to know whether myself and hopefully others should continue working on improving it. Legoktm (talk) 05:17, 28 February 2023 (UTC)
@Legoktm I just responded on Phabricator, but your observations there were correct about the wording. It's unclear whether or not efforts will be made to reconcile API related policies into one page or continue with the different pages, some of which are referenced in the TOU. It seems like many of the API related policy documents are older and need some work to align them with current norms.
For the purposes of the TOU language, we just wanted to be as explicit as possible as to what would constitute a policy so we included the three pages that we did as examples. SSpalding (WMF) (talk) 18:33, 28 February 2023 (UTC)
In addition to the above (not sure it's wise to have a page which explicitly says "This page may be outdated or contain incorrect details" considered part of the TOU), the use of links is confusing. One link is to a whole page, one is to a section, and one is to a section using a method not compatible with all browsers. Should the entirety of the pages be considered part of the TOU, or only the sections linked to? the wub "?!" 16:29, 4 March 2023 (UTC)
Currently, the terms are worded so all policies related to the API, including but not limited to, the ones we linked to are the ones that should be followed. That's a good catch on the confusing hyperlinking. We will make a note to adjust this. SSpalding (WMF) (talk) 00:12, 7 March 2023 (UTC)

API Terms, "...to enable users to build products that promote free knowledge"

I don't see why the API users' intent needs to be stated, my impression we don't really care what people do with the data they get from us, provided they comply with the CC license terms.

Like, if I created a hypothetical app that used the API to pull snippets of factual errors in Wikipedia articles and compared them to a paper encyclopedia in order to disparage Wikipedia and free knowledge rather than promoting it, I think we would be OK with that and not consider it a misuse of our APIs? Legoktm (talk) 05:14, 28 February 2023 (UTC)

I don't think this refers to the users intent, but the intent of those that created the API: "We make available a set of APIs to enable users to build products that promote free knowledge.". If the terms forbade what you could use the API for, that could be considered non-free.
As-is, this seems an empty description. It could as well say: "We make available a set of APIs that enable Weapons of mass destruction to query Wikipedia".
Although not really harmful, I do agree that such part should probably be removed. Moreover, I don't think it is accurate. If I remember right, the API was not made "to enable users to build products that promote free knowledge", but to make the life easier for developers of client/bots (that already existed, but used the more cumbersome GUI), MediaWiki developers (which could then more easily change the interface without breaking bots) and, much more recently, even the software itself (core or extensions), that now use the API, too.
Yurik or Roan might want to provide more historical context.
Platonides (talk) 21:08, 28 February 2023 (UTC)

"all of the content" -> "the vast majority of the encyclopedic content"

Re this:

We do not take an editorial role: Because the Wikimedia Projects are collaboratively edited, the vast majority all of the encyclopedic content that we host is provided by users like yourself, and we do not take an editorial role.

I am concerned by this change. The WMF does not write any part of any of the content pages on the wikis, all is written by volunteers. Why is a change being made that seems to imply that it might no longer be the case? --Yair rand (talk) 07:38, 28 February 2023 (UTC)

Agreed. The impression generated is that some of the encyclopedia content is not generated by users but by the WMF. What is the thinking behind that proposed change? Andreas JN466 14:40, 28 February 2023 (UTC)
I would expect that point of the change is to prevent someone to argue in court that WMF violated the terms because there's some content that a WMF employee contributed somewhere to which someone could refer in court. I think it makes sense to write a Terms of Use with possible court cases in mind, so that change seems fine. ChristianKl ❪✉❫ 14:45, 28 February 2023 (UTC)
But the statement says that content comes from "users" which would include WMF employees who are also users, yes? If taken literally, the current wording does not exclude WMF employees from editing content, but merely means that in doing so, they are not acting in their capacity as employees of WMF. The proposed change does give the impression that WMF intends to contribute, as an organization, to the encyclopedic content. Ashorocetus (talk) 15:59, 28 February 2023 (UTC)
Based on previous suggestions from earlier discussions, the most likely updated language will be "Because the Wikimedia Projects are collaboratively edited, the vast majority of the content that we host is provided by users, and we do not take an editorial role." This omits the word "encyclopedic." Based on previous feedback, we agreed it was unnecessarily confusing since there are projects that aren't encyclopedic but are hosted where we don't take an editorial role. SSpalding (WMF) (talk) 18:41, 28 February 2023 (UTC)
SSpalding (WMF): That leaves the question – What content is not provided by users?
The WMF occasionally uploads its own internal documents, logos, etc., but I would argue that this is not "content" in the sense implied here. I would propose the following wording for 1a:
a. We do not take an editorial role: The Wikimedia Projects are collaboratively edited, and the content we host is provided by users. This means that we generally do not monitor or edit the content of the Project websites, and we do not take any responsibility for this content. ...
That's unless there is a WMF-sourced content category I've missed, in which case could you please give some examples? Regards, Andreas JN466 12:49, 1 March 2023 (UTC)
With the edit ommitting the word "encyclopedic," I hope it's now more clear that we are referring to all content that the Foundation hosts on its servers (which includes Meta wiki). So a good example of content that we edit which has not provided by users is this post here -- which is written by me, a staff member. 🙃 Other examples contest rules for logo contests. It also includes the internal documents, logo, etc that you mention since that is also material that resides on our servers that we host that Foundation staff put there. These are edge cases and we are using the term content broadly. But that's also why the suggested re-wording is currently written as, "the vast majority of the content that we host is provided by uses." SSpalding (WMF) (talk) 15:17, 1 March 2023 (UTC)
Yes, come to think of it, there is MediaWiki as well, for example, where Foundation staff are major contributors obviously.
The Overview at the beginning of the ToU states: Generally we do not contribute, monitor, or delete content (with rare exceptions, such as under policies like these Terms of Use or legal compliance, or when faced with urgent threats of serious harm). This means that editorial control is in the hands of you and your fellow users who create and manage the content.
So given that context I'm inclined to say we're good. Thanks! Andreas JN466 15:45, 1 March 2023 (UTC)
See above, avoid the word "encyclopedic" and don't overcomplicate. Taylor 49 (talk) 08:37, 5 March 2023 (UTC)

What is meant by "false or unverifiable" in section 10?

The change would read "In such cases, we reserve the right, at our sole discretion (or where legally compelled) to... Detect, prevent, or otherwise address fraud, false or unverifiable information security, or technical issues or respond to user support requests;"

Is "false or unverifiable information" meant to stand on its own or is it part of "false or unverifiable information security"? If the former, there should be a comma after "unverifiable". If the latter, I do not know what that would mean.

In the former case, does this give WMF authority to arbitrate what is "false"? How does this square with the principle of verifiability, not truth? Ashorocetus (talk) 16:12, 28 February 2023 (UTC)

Unless "false or unverifiable information security" does have some intended meaning, I would suggest the following to clean up the commas (using an Oxford comma as discussed in #Oxford comma, or lack thereof):
Detect, prevent, or otherwise address fraud, false or unverifiable information, security or technical issues, or respond to user support requests;
On the latter point, it does seem odd to move away from "false or inaccurate" in Section 4, yet include "false" here. But it's not really a big deal to me. the wub "?!" 16:43, 4 March 2023 (UTC)

I also wonder what "false or unverifiable information security" means. As for "false or unverifiable information", that's not appropriate language to include in a binding contract like this, because there's no way to prove what's false or unverifiable. Such a vague and expansive clause exposes users to unlimited legal risk. Did the Wikimedia Foundation commission a legal impact assessment of how "false or unverifiable" is going to play out in jurisdictions around the world? And if not, why was it deemed acceptable to introduce such a legal risk? Nemo 10:15, 31 March 2023 (UTC)

Mer tekst som ikke er oversatt

På Wikipedia på bokmål har vi allerede en lenke til Bruksvilkår som ikke er oversatt, dvs. at den er på engelsk. Nå skal denne allerede lange erklæringen bli enda lengre.

Det er vel og bra, og sikkert svært gode grunner til det, men mitt enkle spørsmål til Wikimedia Foundation er dette: hvordan henger en slik lang tekst, ikke oversatt, sammen med ønsket om å få flere bidragsytere? Dersom det er et krav at man må kunne engelsk for å kunne bidra til Wikipedia, bør ikke det også legges inn i teksten? Ulflarsen (talk) 12:23, 1 March 2023 (UTC)

You point has been heard and noted. We are an international project and each language supports the greater whole, so your critique is valid. Here are some steps WMF has taken or plans to take to promote understandability across languages.
For this terms of use consultation, WMF translated the terms into approximately 8 languages other than English. These languages represent a cross section of the most spoken languages in the world and on-wiki. After the consultation process is over, the final text will likely be translated by the Foundation into around that number as well. From what I understand, this represents more official translations than the existing terms of use. Additionally, all communities are invited to make unofficial translations as well into their own languages of the whole document or just the most important provisions. Finally, machine translations are better than ever (which is one of the reasons I am able to reply to your comment 🙂).
We are open to suggestions though, so if you believe that the machine translation ecosystem is not robust for Norwegian, or that average Norwegians may not be able to avail themselves of other language translations (e.g. German, French, Spanish), or that your community wouldn't have the capacity to create its own unofficial translations into Norwegian in the future, then feel free to let us know. There is always a potential to add new languages to our official translation list if there's an overwhelming need to do so. I'm not familiar enough with Norwegian to know these things, but I am willing to learn. Thanks for your feedback. SSpalding (WMF) (talk) 15:47, 1 March 2023 (UTC)
To SSpalding (WMF): There are of course not "an overwhelming need to do so" (to quote you), as Norwegians managed in many years with encyclopedias on Norwegian, that the ones who could afford it paid for, and the rest could read in libraries. We also have the Great Norwegian Encyclopedia, free of charge. But if we in spite of that think that it could be of use having versions of Wikipedia in Norwegian (that is Nynorsk and Bokmål), then the user terms and related texts, in my opinion, must be in Norwegian. If not, how am I to encourage someone to volunteer as a contributor, when they can not even read such basic documents in their own language? Ulflarsen (talk) 17:00, 1 March 2023 (UTC)
To answer more generally for both you and others who represent Wikipedia's with languages that we may not do official translations for, I would say the best way to do this would be, (1) encourage the potential contributor to use machine translation; or (2) determine if there are any volunteers in your language community who can hand translate either the TOU or just the public explainer document describing the changes; or (3) reach out to WMF movement communications about the problem should neither of these two options be available; or (4) request a grant from the Foundation to subsidize a translation (either encourage the local chapter to do so, or do so as an individual). As an example, there are languages like Catalan where machine translation is still rather buggy, so an option like #3 wouldn't be available there.
Unfortunately, this is not a problem that is unique to this specific TOU updates. Even with the current terms, there are not hand translations for every project language. This something we are attempting to address with the resources we have available to do so. Your feedback helps that. SSpalding (WMF) (talk) 23:13, 1 March 2023 (UTC)

Attribution variants explanation

Please add explanations about possible and wrong ways of how attribution of imported text can be made (dummy edits, face templates, talk templates, etc.). ToU says about sufficient attribution in edit summary, but many projects use talk page templates for attribution (edit summaries could be empty), even without links to concrete edits. Talk page templates can't be used as attribution according to the "(since each article has a history page that lists all authors and editors)" ToU claim with exception if edit summary directly refers to the talk page for attribution. D6194c-1cc (talk) 14:32, 1 March 2023 (UTC)

Also note, that re-using information under CC BY-SA requires to specify whether the text was modified. If someone contributes to the article, page history would have information on every modification of his text. If someone re-uses article text, page history will serve as attribution and as information about modification of every added piece of text. So page history is not only just for attribution. If a dummy edit don't refer to specific edits of pieces of text, then it's almost impossible to determine what and when was modified. D6194c-1cc (talk) 06:44, 4 March 2023 (UTC)

Prohibited use of quotations

In the Proposed update, I don't see addressed the problem of the apparent prohibition of the use of quotations from (non-free) content, not published under a licence compatible with Wikipedia, and the apparent systematic violation of this term on several wikis, including the en-wiki, through the right to quote. (see the item on top of this page and this 2011 article, in Dutch ). --Wickey (talk) 15:51, 1 March 2023 (UTC)

My understanding is that fair-use material is compatible with the CC-BY-SA license (as it doesn't require licensing), and thus quotation is allowed, but perhaps the wording could be clarified to make this more explicit. Nosferattus (talk) 00:19, 3 March 2023 (UTC)
It's covered by the 7 g (Re-use) section. It explicitly says that re-using content that is under fair-use is law-dependent. D6194c-1cc (talk) 13:31, 3 March 2023 (UTC)
Although there is no fair-use term in my county. But quotations are allowed by the law for certain purposes (scientific, academic, information, polemic, educational): [4] (Russian). D6194c-1cc (talk) 13:36, 3 March 2023 (UTC)
As I have indicated, there is obscurity about fair-use. The text should definitely be more clear.
I have never seen on WP a citation from a non-free source with accompanying "relevant attribution ... as they pertain to the underlying license or licenses".
I suggest a separate para about quotations in 7.g and a reference in 7.c. --Wickey (talk) 16:06, 8 March 2023 (UTC)
Although, in my country, laws have separate requirements for the use of illustrations. They can be used for the educational purposes, but not information purposes. So uploading images in many cases is controversial in my country. D6194c-1cc (talk) 09:34, 10 March 2023 (UTC)
I am talking about text quotes. Yet, I see that 7.c. is about text quotes or even pages from sources compatible with CC BY-SA.
Art. 7.g explicitly speaks of re-distribution of wiki pages, or "text content ... imported from another source ... under a compatible CC BY-SA license but not GFDL", or "text that originated from external sources and was imported into a Project ("Project" means WP, Wiktionary, Wikiquote & Co, I presume) that "may be under a license that attaches additional attribution requirements." Yet, I don't see how another "Project" could have imported "text that originated from external sources ... under a license that attaches additional attribution requirements", as that would have been prohibited by art. 7c.
As I understand, for non-text media/images only the laws of the state where the Wikimedia server is applies, that is the USA. Art. 7.g, however, states that the license as stated "on its description page" (that is on commons.wikimedia.org) applies, or "an applicable source page for that work". If the latter is non-compatible with CC BY-SA, the image will not be accepted by Commons in principle, so only "fair use" would be available. Yet, this is forbidden in art. 7d., which requiers "licenses that support the general goal of allowing unrestricted re-use and re-distribution." This would mean that "fair use" is a violation of the terms. --Wickey (talk) 12:20, 12 March 2023 (UTC)

Child sexual abuse

Do pictures like commons:File:Martin Van Maele - La Grande Danse macabre des vifs - 13.jpg violate new variant of ToU? Does this image meet the child sexual abuse criteria? This picture had been proposed for deletion as child pornography, but was kept reasoning that is not a child pornography. I tried to remove this image from Russian wikipedia, but my edit was cancelled (:ru:special:diff/121661878) and administrators didn't see anything wrong with the picture (it was removed from the page afterwards). But this picture is still present in other language Wikipedia projects, and I think it must be removed. If ToU would allow deletion of this image, it would be great. D6194c-1cc (talk) 17:43, 1 March 2023 (UTC)

I'd guess you are referring to the ToU wording
"Posting child pornography or any other content that violates applicable law concerning child pornography or child sexual abuse material;"
The 1905 drawing isn't pornography in any modern sense of the word IMHO. So does this "violate applicable law concerning ... child sexual abuse material"? I don't know - you'd need to know that law very well. But since it is not a realistic drawing, and both people are fully clothed, and the abuse is more hinted at than displayed, I'd think not. Since this might vary by local laws and community sensitivities, it's likely that the matter would have to be decided by local projects by consensus rather than by the ToU. Smallbones (talk) 03:07, 2 March 2023 (UTC)
The men is not fully clothed. As I think this image is sexual perversion with child, and ToU must cover such cases, too. Some related information: en:Wikipedia:Wikipedia Signpost/2023-02-20/Cobwebs, https://web.archive.org/web/20120426201501/https://www.foxnews.com/scitech/2010/05/10/porn-wikipedia-illegal-content-remains/. D6194c-1cc (talk) 08:11, 2 March 2023 (UTC)
The terms of use do cover this - images depicting the sexual abuse of children in a manner contrary to relevant laws are prohibited. Whether a specific image does or does not fall under the definition of "child sexual abuse material" is not a matter for the terms of use - it's also worth noting that different jurisdictions have a wide range of different definitions of what is and isn't child sexual abuse, especially when it comes to drawings rather than photographs. Thryduulf (talk: meta · en.wp · wikidata) 10:07, 2 March 2023 (UTC)
One of the most popular texts on English Wikisource (by pageviews) is The Autobiography of a Flea, an explicitly pornographic 1887 book about a 14-year-old girl who is repeatedly raped by various men. I wonder if it would fall afoul of the new ToU. There may be other material on Wikisource that could be affected as well. If so, this should be brought to their attention before the new ToU goes into effect. For the record, I don't object to such material being deleted (or the new wording being proposed). I just want to make sure this has been considered. Nosferattus (talk) 00:06, 3 March 2023 (UTC)

I'm comfortable with this text

I don't think WMF will ever be able to cover all conceivable conflicts we might get involved in. So neither the current version of the Terms of Use, nor the proposed version, nor any other future version will be found to be perfect. It's likely to be subject to revision from time to time. I'm comfortable with the proposed text. My suggestion would be to adopt this text. My thanks to the people who drafted this text and to the contributors who took time to read and comment on it. MartinD (talk) 08:59, 2 March 2023 (UTC)

Reminder: Office hours about updating the Wikimedia Terms of Use soon!

Hello everyone,

The first office hours with community members about updating the Wikimedia Terms of Use are happening soon!

It will be held in a few hours from now, March 2, from 17:00 UTC to 18:30 UTC. See for more details and the link to the meeting here on Meta.

Another office hours will be held on April 4.

We hereby kindly invite you to participate in the discussion. Please note that this meeting will be held in English language and led by the members of the Wikimedia Foundation Legal Team, who will take and answer your questions. Facilitators from the Movement Strategy and Governance Team will provide the necessary assistance and other meeting-related services.

On behalf of the Wikimedia Foundation Legal Team, DBarthel (WMF) (talk) 09:36, 2 March 2023 (UTC)

Changes to the Paid Editing requirements

I'd love to see us find a method to stop paid editing or - failing that - to control or even moderate its effects. I see it as one of the clearest long-term risks to the projects. The disclosure requirements were an excellent step, but I've encountered two problems.

  • The first is with the gradual increase in restrictions. I support some increases - such as the proposal (not in this version of the ToU) to state that instead of disclosing in edit summaries, on talk pages or on a user page, editors should disclose in edit summaries, talk pages, and user pages, as it does not increase the burden on the disclosed paid editor beyond asking them to do the same thing in three places. But what has worried me is a growing set of rules which make being an undisclosed paid editor far more profitable and easier than being a disclosed paid editor.
  • The second is with the lack of ability to enforce these policies. If a policy is unenforcable, then it is next to useless. We do have the ability to enforce policies that relate to behaviour on-wiki. We can block for lack of disclosure, or for failing to provide a link to off-wiki advertisements, but I have a hard time seeing how we can enforce policies related to off-wiki behaviour beyond legal action.

So given that en.wp has already tried to require paid editors to link to their WP accounts when advertising paid work, and given that a) this was done in part because we were specifically told that it would make it easier to enforce the ToU, and b) that I have now been told that WMF Legal, beyond sending out an undisclosed number of angry letters, have not acted to enforce this because of the cost of legal action, how would making this universal across the projects improve the situation? If WMF Legal have been unable to enforce it when it only applied to en.wp, how will they enforce it when it is universal and there are (presumably) many more cases that they cannot justify paying for legal action against? I gather that the intent is to use the Marketing company mediations to fix this, but if someone is unwilling to abide by the ToU requirements for disclosure, why would they agree to abide by the ToU requirements for mediation? If someone (as I suspect most, if not all, UPE editors will) refuses or simply ignores requests to take part in mediation, and if WMF Legal have been unwilling to pursue legal action against UPE editors in the past due to the cost, why should we expect WMF legal to pursue someone unwilling to engage in mediation when they weren't willing to pursue the same person for failing to disclose?

I am very much at a loss to understand how the mediation improves things, and how expanding rules that WMF Legal have been unwilling to enforce via legal action to date, helps solve a very real problem. - Bilby (talk) 04:58, 3 March 2023 (UTC)

We all know the problem; but is there any solution without changing the principle that contributions may be anonymous? I used to think perhaps we could ask for identification only for articles about commercial businesses and their officers, but paid COI can be relevant to almost any type of article. I thought we could perhaps keep the verification confidential, but this would either place an impossible burden on the volunteers, or require using paid staff (and I certainly do not think it wise to add any additional paid staff for any purpose) Similarly I used to think that we could help by raising the notability standard for businesses, but we have in effect done that--and it has helped--butt that still leaves all the other types of possible articles (for the material I work with, the worst offenders are non-profits) DGG (talk) 20:48, 3 March 2023 (UTC)
We could be willing to legally enforce the ToU as it stands rather than introduce things that will just be ignored. Or we could open up CU to do regular checks for known paid editors. And I agree that narrowing the requirements for businesses helped - I'd like to see us do the same for BLPs, so every second plastic surgeon stops trying to get an article. - Bilby (talk) 21:09, 3 March 2023 (UTC)
I'd like to address some of your points here, but for anyone else reading, I want to call attention to the section Marketing Company Mediations where many points addressed here are dealt with in greater detail.
I agree that extra rules have a tendency to complicate things for good-faith users while bad-faith users just ignore them. This is why the TOU update does not create any meaningful new rules. I also agree that we should aim legally enforce the TOU as it stands. This is why rather than introducing new rules, we are introducing a streamlined mechanism to enforce the existing rules. That way we can enforce the rules more often and more systematically.
You asked, "how [does] expanding rules that WMF Legal have been unwilling to enforce via legal action to date, helps solve a very real problem" To answer your question as directly as possible, we have tried to enforce these rules over the years with limited success. We are very willing to do this. We have successfully stopped companies in the past. But learning from the limitations we've faced, we have suggested this new process help us do this better. The details of what that means are described in much more detail in that section, but I think something important to note is that (1) community rules are different than (2) the TOU. Editors will still enforce on-project UPE rules in whatever manner is deemed appropriate. This is creating a new, potential tool for the Foundation dealing with activity off-wiki.
You are right that companies would not likely voluntarily want to submit to mediation. Mediation/arbitration is enforced through court orders. Getting a court order to force an entity into mediation could be a small fraction of the cost of litigating against them. Getting a court order to enforce the outcome of the mediation would also be a small fraction of the cost of getting a court to enforce a judicial judgement. Currently, even if we won in litigation, we would have potentially hundreds of thousands of dollars in attorneys fees to get a judgment that could also cost more money to enforce. Finally, in mediation, we would be able to get a court to order that the company we litigated against repay our attorneys fees. These are some of the differences which make mediation both more threatening to these companies (making letters more useful) and make it more likely that we will take enforcement actions (because it's both less expensive and there may be a reasonable chance we can recoup our lost attorney's fees). Again, this and other benefits are discussed with a lot more nuance in Marketing Company Mediations for others who may be reading about this here for the first time.
From your feedback, it seems like your bigger picture concern is more generally that you do not believe that this will work. That is reasonable. I can't make any guarantees that this will work either. What I can say is that this update was drafted specifically to work again a handful of specific companies we already have an interest in enforcing against who we otherwise wouldn't have the tools to deal with.
With this addition, our intent is to use what we've learned over several years of enforcement actions to try to improve our process which we feel is better than accepting the status quo. My goal during this consultation is to give everyone the most transparent description of our intent behind the changes possible. I think it's reasonable for people to have a wide range of opinions on whether or not the changes will be effective, and it's totally ok to see this as a fool's errand.
I appreciate you bringing your experience as a long-term volunteer to this. I'm sure you are not the only person in the community who feels this way. Should this language be welcomed by the community such that we can try to use mediation against UPE companies, then your feedback has motivated me to show -- through actions rather than just words -- that this can be effective in certain situations. One nice thing is that since on-wiki project rules are not meaningfully changing here, that means the most likely scenario in the event this fails is that no-one contributing to the projects even notices that anything changed. 🙃 SSpalding (WMF) (talk) 23:25, 3 March 2023 (UTC)
I understand this, and I did have in mind that a court order is cheaper than litigation. And ultimately, I don't suppose you need the community's support. However, I have been fight paid editing for as long as anyone on wiki, and I've come to accept that we have a situation where the methods of stopping it have to either revolve around technology to identify paid editing; changes in policy that reduce the demand for paid editing; and changes in policy to make it easier to act on-wiki against paid editors. The problem is the amount of money they make and the impact on their lives. At the moment, I have a list of 285 paid editors identified through Upwork, Guru and the other freelancing sites. Flicking through some of the accounts, I can see people are making tens of thousands of dollars a year through paid editing, and they are working from Bangladesh or Pakistan Where I'd be concerned that enforcing the court orders is going to be difficult at best. We are trying to take away their livihoods, and I can't see them willingly accepting that when they can just ignore court orders and place the burden on the WMF to try and have them enforced. However, yes, I agree that the worst case scenario for the changes to the ToU will be that it fails and doesn't directly impact editing, however it also, if it fails, will act to divert attention from methods that might work. So would the WMF agree to provide a report on the success of this in, say, 12 months? How many companies or individuals stopped paid editing as a result of WMF action? How many times it went to mediation and it was successful? Because if it continues to look like there are effective tools in place, even when they aren't effective, we're not going to be making changes that matter. - Bilby (talk) 00:20, 4 March 2023 (UTC)
I'll try to make @Bilby:'s example even more concrete and give another concrete example from today (leaving out the company's name). Is it likely that the new ToU will lead to enforcement in the following cases:
You get a list of editors who use Upwork or other companies who don't link back to a paid editing declaration on a Wikipedia page. Can you take Upwork or the other companies to mediation if they don't kick off the editors from their sites?
An article about a paid editing company appears in the press (likely paid coverage) which links to their website which lays out in some detail how they create and maintain Wikipedia articles. Specifically they say that 1) they follow all Wikipedia rules, 2) that the client's name will never be released, and 3) that they've edited and maintained hundreds of Wikipedia articles over the last 5 years. Add in 4) that there are no paid editing declarations on-Wiki that mention the paid editing company's name.
Would you be able to get these folks into mediation? If not, what else do you need? Can you describe a "perfect case" that almost certainly would be taken to mediation, and what evidence we'd need to provide? I'm not really trying to put you in a corner, just trying to figure out what amount of evidence I'll need to collect. But if, after I collect the evidence needed, you tell me I need a photocopy of the check used to pay the editor, or something as impossible as that, I will be disappointed Smallbones (talk) 04:40, 4 March 2023 (UTC)
No worries. Both you (@Smallbones) and @Bilby have spent a great deal of time and energy helping combat UPE on the projects, so the least I can do in return is to try to explain what we are trying to do in a comprehensive way. There are several common threads running through your replies, so I'll start with
1. What would make a perfect case?
I know this is not a satisfying answer, but if I were to describe the perfect case, then it would be easy for any company that hits those criteria to avoid doing those things and do something "slightly different" that makes them a more middling target. Let me elaborate with an example of a "good enough" case that I'm more comfortable describing.
Imagine an Upwork profile that only makes money through editing Wikipedia, that we know is part of a network because [REDACTED], the firm is clearly active, we can prove that they are making a large amount of money because [REDACTED], the company advertises [REDACTED] on its profile that we see as an obvious TOU violation on its face, and we can connect this activity to a larger business because [REDACTED], and we know that the company exists in a jurisdiction with a predictable judicial system because [REDACTED]. We have contact information because [REDACTED].
Even with the redactions, it's difficult to elaborate. As soon as I lay out criteria like this, it's easy enough for a firm to change some of these business practices. For example, the owner of the profiles could fake it so it looks like they make money through other sources other than editing Wikipedia. They could do things to suggest that they're no longer active, etc. Said differently, describing who is or isn't a good target and how we go about figuring that out would give that company an easy route to cover their tracks. The difference between a company being a perfect target and being a middling target might be a handful of small decisions that the company makes.
My fear is that since we currently have a small handful "good" targets, if we lose one or several of them because a company representative read this answer and slightly changed their business model, then I would consider that an enormous loss. I know that there was some earlier dismissal of the position that the "WMF can't discuss certain things because of confidentiality" position, but that is real. Let me elaborate on that below.
Some of the firms that this would be most likely used against already know they are targets because we have active or semi-active disputes with them. These companies don't want to lose $100,000's per year. If they are represented by anyone who is competent, that lawyer would start their defense by reading the Foundation's own statements we make on this discussion page and use them against us (either for clues to evade detection or as "statements against interest"). To give a more nuanced answer to an earlier question about why even the nature of letters we send is confidential: a company with lots of money at stake will view risk much differently if the Foundation is sending 50 toothless letters a week versus if the Foundation only sends out a handful of targeted letters every quarter. In other words, if I know I'm one of 10000 being targeted, then it's much different than if I know that I'm one of 10 being targeted.
With that background in place. To directly answer some of @Smallbones 's questions
A. The TOU would apply to companies that use the Wikimedia projects (since they would have agreed to the Wikimedia TOU), not to Upwork as an entity since Upwork hasn't agreed to our terms of use. That doesn't stop us from working alongside Upwork to get rid of certain profiles but we wouldn't "mediate against Upwork"
B. If someone makes "statements against interest" that suggest they violate the rules, that could make them a good target. I doubt that we'd take a statement that says "we follow the rules" at face value when other evidence would suggest something contrary.
2. Can we provide a report/year in review?
Your feedback has reinforced the need to ensure everyone sees that something meaningful comes from this. But for all the reasons above, there would be some significant guardrails on providing information on our strategy. To the extent that we have to file court orders, then those would be in the public record. I imagine we'd be able to give some public information about the disposition of that, but I imagine information about our strategy wouldn't be disclosed after the dispute itself is resolved (or possibly long after). To explain our legal strategy publicly on-Wiki, would require us to provide information that could be useful to the other side to avoid enforcement. Said differently, once we develop a winning strategy, it would make much more sense to "repeat that winning strategy" on new similarly situated targets rather than disclose the nuances of the strategy and allow targets to adapt.
Also, I imagine that the pace of this will be rather slow, so 12 months from now it may be reasonable that few actions may have escalated to the point that meaningful public information can be revealed (see above).
Your big picture point, that we should attempt to do things that will have a meaningful effect, is well noted. I think it's worth also setting expectations here that UPE will always exist regardless of any efforts to deter these companies from wasting volunteers' time. What we're attempting to do is raise the cost of doing business for the companies such that it increasingly makes less financial sense to be in the business of coordinating UPE forcing them to move on to some other more lucrative pursuit. I think that this change would definitely affect the costs/risks for these companies, so I'd consider this change a meaningful one.
3. What extra evidence do editors who want to help out with this need to collect?
One nice thing is that the current systems of surfacing problematic accounts, blocking them, and discussing why they are problematic on Wiki is sufficient to power this process. To the extent that we need more information, we would likely just ask for what we needed. SSpalding (WMF) (talk) 00:06, 7 March 2023 (UTC)
I very much appreciate you taking the time to reply. As I understand it, then, this change is unlikely to have any significant effect on paid editing in the next 12 months, and is largely targeted at a small number of companies, so it won't be able to be used against companies such as Upwork or, in all likelihood, the majority of people engaged in paid editing. And, as with the previous change on en.wp requiring people to link to their wp accounts when advertising, we won't really be able to find out anything about the effectiveness of these changes, except maybe anecdotally through what we can see, as it is not possible for the WMF to provide a meaningful report on how these changes have assisted in combatting paid editing because of confidentiality. But given that these changes are limited in who they can be used against, we're unlikely to see anything of significance anyway, so if - as an example - the en.wp community wishes to more aggresively tackle paid editing, they will need to do it through internal change, and won't see much through the WMF. Which is pretty much the status quo anyway. - Bilby (talk) 03:42, 8 March 2023 (UTC)
To clarify some points.
  1. we believe that this will have an immediate effect upon implementation. The mere fact that the rules are more likely to be enforced makes the most organized and systematic players in the UPE space more careful and less prolific even without enforcement actions happening within the first 6 months. What I meant was that expectations should be tempered: all efforts against UPE simply manage the problem as long as the profit motivation to distort and bias projects is there. There are individuals, especially smaller firms that are not the most ideal targets, that will never care about rules.
  2. It will not be used against Upwork the company, but it's very likely to be used against the firms that advertise their services through Upwork profiles. Upwork itself is not bound by our TOU, but companies that edit the projects that advertise on Upwork are.
  3. I think that's fair to say that effectiveness reporting will be anecdotal. I would say that we have identified that a small amount of bad actors create an outsized amount of UPE so frustrating those actors could be significant. But there's no guarantee that new actors don't take their place just as quickly.
Thanks for all of the useful feedback and all the work you do for the projects!
SSpalding (WMF) (talk) 18:24, 8 March 2023 (UTC)

Nutzungsbedingungen in Deutsch / Terms of Use in German

Im ersten Abschnitt wird über die rechtliche Situation der Wikipedia in Amerika geschrieben

"[...] ist eine 501(c)(3) nichtkommerzielle wohltätige Organisation mit Sitz in San Francisco, Kalifornien [...]"

Deutsche, bzw. deutschsprachige Mitglieder können mit der Angabe eines Gesetzesabschnittes eines US.-Amerikanischen-Steuergesetzes nicht viel Anfangen. Ich habe gerade etwas suchen müssen um heraus zu finden, dass es sich dabei um den USC 26 §501 handelt.

"[...] ist eine nichtkommerzielle Wohltätigkeitsorganisation mit Sitz in San Francisco, Kalifornien [...]"

würde mMn sich deutlichst besser lesen. Wenn der Hinweis auf die steuerliche Situation zwingend notwendig ist

"[...] ist nach USC 26 §501(c)(3), eine nichtkommerzielle Wohltätigkeitsorganisation mit Sitz in San Francisco, Kalifornien [...]"

grüße, Adtonko (talk) 13:35, 4 March 2023 (UTC)

In the first paragraph, it is written about the legal situation of Wikipedia in America

"[...] is a 501(c)(3) non-commercial charitable organization based in San Francisco, California [...]"

Germans, or German-speaking members can't do much with the specification of a law section of a U.S.-American tax code. I just had to do some research to find that it is USC 26 §501.

"[...] is a nonprofit charitable organization located in San Francisco, California [...]"

would read significantly better in my opinion. If the reference to the tax situation is mandatory

"[...] is, under USC 26 §501(c)(3), a noncommercial charitable organization located in San Francisco, California [...]"

greetings, Adtonko (talk) 13:35, 4 March 2023 (UTC), translated by DBarthel (WMF) (talk) 15:38, 5 March 2023 (UTC)

Hej Dennis, you forgot an 's' on the 'Germans' - i can be nitpicky about my own spelling 'scuhen' or 'heruas' ... anyways, thanks for translating!
- Adtonko (talk) 18:46, 6 March 2023 (UTC)
Fixed :) DBarthel (WMF) (talk) 12:04, 7 March 2023 (UTC)

A few copyedits

I have a few copyedits that I'd like to suggest:

  1. "Click “help” on the left side of most pages" – replace this with "Click “help” on the sidebar of most pages". It's more concise and less vague.
  2. "headquartered in San Francisco, California" – replace this with "headquartered in San Francisco, California, United States". California is a subnational jurisdiction; although it's widely known in most places, we still cannot expect everyone to know where this subnational jurisdiction is (and it's no different to saying, "Vitória, Espírito Santo", per se
  3. in § Overview – why is "project" capitalised? It is not being used as a proper noun here.

That's all I have for now. I may add a few more when I stumble across them. Otherwise, I'm all for these new updates :-).

--SHB2000 (talk | contribs) 23:01, 4 March 2023 (UTC)

Deutsche Übersetzung 2023 / German translation 2023

https://meta.wikimedia.org/wiki/Wikimedia_Foundation_Legal_department/2023_ToU_updates/Proposed_update/de sagt:

Zu unseren Zielen gehört es:
Wir bieten Websites und technische Infrastruktur an, um Ihnen dabei zu helfen.

Da stimmt die Grammatik nicht.

501(c)(3)

Für deutsche Leser ist nicht klar, was das bedeutet. AGB müssen klar und allgemeinverständlich formuliert sein.

Wir heißen Sie („Sie“ bzw. den „Benutzer“)

Das wirkt wie eine Entwurfsfassung.

oder Beitragender

oder Beitragenden

Universal Code of Conduct (UCoC)

Dieser Begriff wird nur 2-mal erwähnt, daher lohnt sich die Einführung einer Abkürzung nicht.

Partizipation

Teilnahme

Verletzung des nach dem Recht der Vereinigten Staaten von Amerika oder anderen anwendbaren Rechtsordnungen (darunter gegebenenfalls derjenigen des Landes, in dem Sie leben oder Inhalte betrachten bzw. bearbeiten) bestehenden Rechts auf Privatsphäre anderer;

Der Satz ist zu verschachtelt, um ihn mit zumutbarem Aufwand zu verstehen.

Person oder Einrichtung … Personen oder Einheiten

Ist mit Einrichtung und Einheit dasselbe gemeint?

Missbrauch unserer Dienste zu anderen illegalen Zwecken
* Einstellen von Kinderpornografie oder von anderen Inhalten, die gegen geltendes Recht bezüglich Kinderpornografie verstoßen oder Materialien des sexuellen Missbrauchs von Kindern;
* Einstellen oder Austausch von obszönem Material, das unter geltendem Recht rechtswidrig ist; sowie die
* Nutzung der Dienste in einer Weise, die nicht mit geltendem Recht vereinbar ist.

Fällt der erste Punkt nicht sowieso unter den dritten Punkt und ist damit redundant?

oder Materialien des sexuellen Missbrauchs von Kindern

Ist ein einzelnes Material OK? Oder warum die Formulierung "Materialien"?

Wie in Abschnitt 14 dieser Nutzungsbedingungen beschrieben.

Dieser Satz kein Verb.

Auch wenn Ihnen hinsichtlich der Weiternutzung von Inhalten auf den Projekt-Webseiten beträchtliche Freiheiten gewährt werden, ist es wichtig, dass wir als Wikimedia Foundation unsere Markenrechte schützen, um dadurch unsere Benutzer vor Nachahmern zu schützen, die betrügerische Absichten verfolgen. Aus diesem Grund möchten wir Sie bitten, unsere Markenrechte zu respektieren. Sämtliche Marken der Wikimedia Foundation sind Eigentum der Wikimedia Foundation und jegliche Nutzung unserer Handelsmarken, Marken, Dienstleistungsmarken, Logos oder Domänennamen muss in Einklang mit diesen Nutzungsbedingungen sowie in Einklang mit unserer Richtlinie betreffend Markenzeichen erfolgen.

Geht das auch etwas direkter, statt um den heißen Brei herumzureden? Wie wär's mit "Wenn Sie eine unserer Handelsmarken, Marken, Dienstleistungsmarken, Logos oder Domänennamen nutzen wollen, lesen Sie die Richtlinie und kontaktieren uns in Zweifelsfällen."

Sie [hier] weitere Informationen

Das Wort "hier" ist nicht verlinkt.

--RolandIllig (talk) 14:44, 5 March 2023 (UTC)

https://meta.wikimedia.org/wiki/Wikimedia_Foundation_Legal_department/2023_ToU_updates/Proposed_update says:
: Our goals include:
We provide websites and technical infrastructure to help you do that.
There's grammar wrong.
: 501(c)(3)
For German readers it is not clear what this means. Terms of Use must be worded clearly and in a way that is generally understandable.
: We welcome you ("you" or the "user").
This sounds like a draft version.
: Universal Code of Conduct (UCoC)
This term is only mentioned 2 times, so it is not worth introducing an abbreviation.
: Violation of the privacy rights of others under the laws of the United States of America or other applicable jurisdictions (including, if applicable, that of the country in which you live or view/edit content);
The sentence is too complicated to be understood with reasonable effort.
: Misuse of our services for other illegal purposes.
* Posting child pornography or other content that violates applicable law regarding child pornography or child sexual abuse materials;
* Posting or sharing obscene material that is illegal under applicable law; and.
* Use of the Services in a manner inconsistent with applicable law.
Doesn't the first point fall under the third point anyway and is therefore redundant?
: As described in section 14 of these terms of use.
This sentence is missing a verb.
: Although you are granted considerable freedom with respect to your continued use of content on the project websites, it is important that we, as the Wikimedia Foundation, protect our trademark rights, thereby protecting our users from copycats who may have fraudulent intentions. For this reason, we ask that you respect our trademark rights. All Wikimedia Foundation trademarks are owned by the Wikimedia Foundation, and any use of our trademarks, brands, service marks, logos, or domain names must be in accordance with these Terms of Use and our Trademark Policy.
Instead of beating around the bush, can you be a little more direct? How about "If you want to use any of our trademarks, brands, service marks, logos or domain names, read the policy and contact us if in doubt."
--RolandIllig (talk) 14:44, 5 March 2023 (UTC), translated (not including feedback on the German translation) by DBarthel (WMF) (talk) 13:49, 6 March 2023 (UTC)
@RolandIllig - danke für die Hinweise. Die Punkte, die direkt mit der deutschen Übersetzung zu tun haben, sollten behoben sein. Ich habe sie daher nicht in die Übersetzung aufgenommen. / Thanks for the hints. Points directly related to the German translation should be fixed now. I have therefore not included them in the translation. DBarthel (WMF) (talk) 13:55, 6 March 2023 (UTC)

Illegal content

The new TOU would state You do not...post illegal content, but there's a glaring omission there. "Illegal" under the laws of what jurisdiction? For example, I am quite certain that much of the English Wikipedia article about Tiananmen Square is "illegal" under Chinese law, and the article about Kim Jong-Un is almost certainly "illegal" under North Korean law. I don't think the TOU is meant to prohibit those things, but without specifying a particular jurisdiction, it would. Seraphimblade (talk) 16:19, 6 March 2023 (UTC)

@Seraphimblade see the #Lawful Behaviour section above where there is existing discussion about this. Thryduulf (talk: meta · en.wp · wikidata) 01:38, 7 March 2023 (UTC)
Thryduulf, it's actually #Lawful Behavior, but thanks for the pointer in any case. Seraphimblade (talk) 02:04, 7 March 2023 (UTC)

Using Wikimedia projects for expression of beliefs and views that cause harm

In the light of legal cases in the last few years, and changes in legislation under US law and the law in other countries for "belief" or religious views becoming understood as protected characteristics, how does the Terms of Use reconcile the protection of people holding beliefs and religious views, and does the ToU act as a double-edged sword for minority communities that are actively harmed by the expression of extreme views? My reading is that under the ToU and UCoC, expressing "extreme" beliefs is allowed, only if they are expressed in a way to appear like a personal attack would they be considered unacceptable conduct. Clearly this can lead to gaming the system, with someone being offensive and causing distress, but claiming they are the one being harassed if anyone challenges them on their views or asks them to remove their opinion.

I have in mind users who may (claim to) feel they have a right to express their beliefs and views as opinions in discussion and formal requests for comments, even though one interpretation of the ToU makes those views unacceptable and sanction-able. For example views that deny the existence of transgender people, or promote conspiracy theories about the Covid-19 pandemic or historical accounts of the holocaust.

Though there may be local policies on the English Wikipedia for some of these topics, there are many other projects where no direct policy exists, for example on Wikimedia Commons it may be entirely valid to host extremist propaganda and discuss that propaganda, even though these may harm and distress minority communities, including Wikimedia volunteers, which they attack or defame. Anstil (talk) 10:30, 7 March 2023 (UTC)

inappropriate sentence

this : "Publier ou modifier du contenu dans l’intention de nuire gravement aux autres, comme des incitations délibérées à l’automutilation ou le déclenchement délibéré de crises d’épilepsie. " should be only : "Publier ou modifier du contenu dans l’intention de nuire gravement aux autres."

Post or edit content with the intent to cause serious harm to others, such as deliberately inciting self-harm or deliberately inducing seizures. should be only: Post or edit content with the intent to cause serious harm to others. Domi1789 (talk) 09:39, 8 March 2023 (UTC)

@Domi1789:, Je ne suis pas wmf, juste un autre utilisateur comme toi, mais quelle est ton objection à la langue existante ? Je ne vois aucune raison de supprimer quelques exemples. Il me semble que l'utilisation du mot "délibéré" dans les exemples est essentielle. Par exemple, l'image en haut de en:Trypophobia (voir la PdD si tu ne comprends pas pourquoi cette image est problématique pour certaines personnes). Note aussi que fr:Trypophobie évite cette image ; peut-être exprés ? Mathglot (talk) 20:41, 8 March 2023 (UTC)

Suggest dropping the "Universal code of conduct" addition

The "Universal code of conduct" is controversal....partly due to problematic inclusions, and partly due to having been developed in an ivory tower rather than by the community. Suggest dropping adding it to the terms of use. North8000 (talk) 14:58, 8 March 2023 (UTC)

Seconded. Dahn (talk) 09:10, 9 March 2023 (UTC)
+1 Hund96 (talk) 08:33, 15 March 2023 (UTC)
Sorry but that is just ignoring reality. The UCoC is in effect since it has been approved by the Board of Trustees in December 2020 [5][6] per current Terms of Use, Section 11 [7] -> binding resolutions by the Board of Trustees. Therefore it is simply the right thing to explicitly mention the UCoC in the ToC.
What's been controversial were the enforcement guidelines. But as the guidelines received 76% support in the recent vote [8] with 3097 users from 146 Wikimedia communities voting it's time to accept that they are going to be part of the policy as well. By the way: These guidelines were developed by a committee largely consisting of volunteers from various local wikis [9] and not „in an ivory tower rather than by the community“. Johannnes89 (talk) 09:05, 15 March 2023 (UTC)
It was suggested the final draft of the guidelines that: [the] UCoC be added to Wikimedia Terms of Use.
Previously, it stated that Users with enhanced rights such as, but not limited to: sysop, bureaucrat, steward, interface admin, checkuser; should be required to affirm (through signed declaration or other format to be decided) they will respect and adhere to the Universal Code of Conduct.
I much prefer simply including the UCoC in the ToU as recommended in the final draft. –MJL ‐Talk‐☖ 06:28, 12 April 2023 (UTC)
I agree. Vecr (talk) 22:55, 8 April 2023 (UTC)

Mission statement

Mission statement > Under..

> Lawful..
what's "illegal content"? What country? For example: en-wiki is uncensored, and has plenty of explicit sexual imagery, which I can imagine might be illegal in some countries. Political example: calling the Russia-Ukraine war a "war" is illegal in at least one country. Religious: images of Muhammad are illegal in some countries. Copyright: varies from one country to another (maybe you've already dealt with this in the copyright section; haven't got that far yet).
> No harm:
The structure "You do not AAA and BBB" is semantically ambiguous in English, because you could parse it logically as "you ([do] NOT AAA) AND ([do] BBB)" or "you [do] NOT ([do] AAA AND [do] BBB)" which is different. The problem is, that it's not clear how to parse the expression following the NOT. If you flip the order so that the NOT can only apply to the second one, the ambiguity is removed:
... you follow the policies for our technology infrastructure, and do not harm it.

That's as far as I got for now. Mathglot (talk) 20:16, 8 March 2023 (UTC)

New terms of use is way better

Its clarity helps keep vandals off the platform. No more vandals :) Paintkitty3 (talk) 06:29, 9 March 2023 (UTC)

"Be part of a global community that will avoid bias and prejudice"

As we stand, the Code of Conduct includes this: "all who participate in Wikimedia projects and spaces will [...] [b]e part of a global community that will avoid bias and prejudice". That's all fine on the surface, however, I worry that it may be construed as a carte blanche to remove or tweak article content that is about people who held prejudiced opinions, or have been said to hold such opinions. For instance, while our rule is to give due coverage to one's antisemitic opinions in an article about a notable antisemite, without implying that the subject was right to disseminate such opinions, and where we would qualify the opinions by citing a secondary, reliable source expressing a qualified view as to why the opinion is stupid, I can easily see enthusiastic censors feeling validated by the Code of Conduct and proceeding to simply remove the opinion in its original expression. Dahn (talk) 09:19, 9 March 2023 (UTC)

Wording of message

The message "The Wikimedia Foundation is updating our Terms of Use", displayed at Wikimedia Commons (and I suppose probably elsewhere), reads rather strangely, as if the Wikimedia Foundation is changing Wikimedia Commons' terms of use (or the terms of use of whatever site it is displayed at). In other words, "our" seems to refer to the site on which the message appears, not the Wikimedia Foundation. I wonder if this is actually what was meant. ITookSomePhotos (talk) 22:12, 9 March 2023 (UTC)

The Wikimedia Terms of Use applies to all projects. In other words, the Wikimedia Terms of Use would apply to you if you use Wikimedia Commons, Wikinews, or Wikipedia. Although Wikimedia Commons does have unique on-project rules (that this update will not change), Wikimedia Commons does not have its own separate terms of use. That is why these terms of use changes are being advertised across projects and discussed on MetaWiki. SSpalding (WMF) (talk) 22:53, 9 March 2023 (UTC)
Thanks, but I don't think you have understood my point. Either that, or you have confirmed that the message is indeed misworded. ITookSomePhotos (talk) 23:00, 9 March 2023 (UTC)
@SSpalding (WMF) I think they meant they'd read it as that "our" can read as "The WMF's TOU", rather than "all Wikimedia projects' TOU". Nosebagbear (talk) 09:43, 10 March 2023 (UTC)
Understood. I agree, that is confusing. Duly noted SSpalding (WMF) (talk) 14:27, 10 March 2023 (UTC)

Good things

There is always a tendency for "feedback requests" to turn into an airing of grievances. To buck that trend, I want to highly that I really like the new paid editing provisions (especially Med-Arb) and the license upgrade. Thanks for your hard work in making this happen! HouseBlaster (talk) 02:41, 11 March 2023 (UTC)

Updates to the updates

Question: are we going to see proposed updates to these updates in response to feedback during this process, or will that happen afterward? Thanks, HouseBlaster (talk) 02:41, 11 March 2023 (UTC)

@HouseBlaster: please see below for an update based on the feedback collected in the first month of the cycle. Thanks! RamzyM (WMF) (talk) 12:20, 22 March 2023 (UTC)

Don't condition Readers' rights on this ton of legalese

People who merely read Wikipedia projects should not have to accept a mandatory Code of Conduct (that says nothing about the conduct of reading), and 99% of the rest of this "Terms of Use". Another way to say it is that these long, highly detailed and prescriptive terms should only apply to the tens of thousands of editors, not to the hundreds of millions of readers. Gnuish (talk) 09:55, 11 March 2023 (UTC)

Maybe that's the point. 2dk (talk) 16:35, 11 March 2023 (UTC)
What is the point of forcing a reader to accept this contract of adhesion? None. The UCoC and its Enforcement Guidelines were completely written from the point of view of controlling the behavior of people who contribute to the project -- those who write or edit text, who leave talk page comments, who attend a conference, who interact with other users. That was its purpose and that is who it applies to. Yet there are literally 1000+ READERS for every EDITOR. We should not drive that thousand away by telling them that they need to read and follow a bunch of legalese that simply doesn't apply to them.
The UCoC says "The Universal Code of Conduct applies equally to all Wikimedians without any exceptions", but it never defines what a "Wikimedian" is. The rest of the text refers to people who write, or who come to a Wikipedia conference or other physical event; for example in the sentence "Every Wikimedian, whether they are a new or experienced editor, a community functionary, an affiliate or Wikimedia Foundation board member or employee, is responsible for their own behaviour." Nowhere in that list (or anywhere else in the UCoC) does it include "reader". I can find no part of the UCoC that might apply to someone who comes to Wikipedia and reads two or three pages and then goes elsewhere in the web. How can such a person "show respect for others" -- by which article links they click on? By how long they stay on one page before reading the next page? How could a reader use "Insults" or "Harassment" in how they read Wikipedia? The only thing that they can type goes into the search box and is never seen by any other human.
If you think the UCoC should apply to people who merely read on Wikipedia, please identify what section or what text in the UCoC you think should apply to them -- and how you would propose to sanction an anonymous reader for violating that text. Gnuish (talk) 04:13, 16 March 2023 (UTC)

Licensing of scripts and other code posted on-wiki

Code created by community members, such as Gadgets, Python scripts, JavaScript code or CSS code is often posted on-wiki, and intended either for personal or general project-wide use. Nowadays not only the usual JS/CSS exists, but Lua modules are available too, which have become increasingly important for the wikis functionality. Some Gadgets and Lua modules are important and complex pieces of code.

Section 7 of the proposed Terms of Use (PToU) seems to remain silent about this type of content, focusing only on text and non-text media. I guess it all boils down to the question if code is legally text or non-text media. Unless there's legislation or case law regarding that, I think code does not really fall cleanly in either category. Code is only "text" (PToU § 7(a)) inasmuch it has to be written down and is being published in a page. It doesn't feel like "non-text media" (PToU § 7(d)) either, as that paragraph read in combination with the wmf:Resolution:Licensing policy cited is obviously focused on multimedia content (images, videos, etc.) only.

Now, while re-using said codes inside Wikimedia projects seems well covered by Section 7(g) and (h), it looks to me external re-use and adaptation of such codes for e.g. bots or other non-WMF wikis may be problematic.

While Section 1(a)(h) of the CC-BY-SA 4.0 license defines «Licensed Material» as «the artistic or literary work, database, or other material to which the Licensor applied this Public License» (and that «other material» could be code), Creative Commons strongly advice against using CC licenses for software for a number of reasons. In particular, their licenses lack «specific terms about the distribution of source code». They also lack specific terms regarding patent rights, and are «not compatible with the major software licenses». While CC 4.0 has a one-way compatibility with the GNU Public License v3 it is not automatic and reserved only for specific situations. Similarly the GNU Free Documentation License is mostly focused on distributing text-based works and recommend using their General Public Licenses for non-text contributions. According to Creative Commons, currently, no non-CC licenses have been designated as compatible with BY-SA 3.0. As such external reusers will be forced to stick with a license that is not specific for software.

I suggest to either clarify what the WMF considers Gadgets, JavaScript, CSS code and Lua modules, etc. to be; or add a new paragraph to cover those including other executable code posted on-wiki. I also think we should be explicitly allowing authors to release such codes under the terms of a Open Source Initiative approved license (OSI) of their choice or, in the alternative (absent a choice or where the choice is void), the Terms of Use should mention a default license for such content under an appropriate software license, so all submitted content «is freely reusable by anyone who cares to access it» as the Terms say; with an appropriate software license.

Please note that wmf:Policy:Terms of Use/Phabricator exists, which deals specifically with code submitted to Wikimedia Phabricator (curiously, Wikimedia Gerrit, where most MediaWiki development happens, is not covered).

Thank you, —MarcoAurelio (talk) 20:36, 11 March 2023 (UTC)

@RamzyM (WMF): Hello. Could you please flag this to the team for review? Thanks, —MarcoAurelio (talk) 12:46, 22 March 2023 (UTC)

Hi MarcoAurelio, RamzyM flagged this to us and I can offer some answer here. I'd divide this into a few different ways that code could get on the projects.

  1. Code someone wrote that's already licensed elsewhere (like Github or a similar repository): The license from that other location would still be the primary license for the code and generally those licenses require that they be kept with the code, so someone posting it on-wiki would need to clearly specify that it's, say, Apache 2.0 licensed code. Putting it on-wiki in this case probably would also apply a CC BY-SA license to the text on-wiki, which adds very little value to people running the code but does add value for people re-using Wikipedia in bulk by having fewer attribution exceptions. For example, if I want to just grab a whole series of pages and reproduce them for a presentation, I can do that under CC BY-SA and not have to worry about the embedded code on one of those images not being licensed the same way.
  2. Code that comes from a third party source: In this case, the person reproducing the code probably doesn't have the right to license it under CC BY-SA. They can still post it on-wiki as long as they follow the terms of the original license (again using my Apache 2.0 example, they just need to keep the license notice with the code they're posting). While this scenario isn't described under section 7 of the ToU, it's completely fine and doesn't violate any copyright laws. There is an extremely small chance of a third party re-user of Wikipedia somehow copying part of the code but snipping out the original license thinking they're allowed to do so under CC BY-SA 4.0. This is, honestly, so obscure and low risk that I think it would not be worthwhile to do any legal work to try and cover the scenario under our ToU.
  3. Original code posted only on-wiki: In this case, the code would be owned by the poster as all-rights reserved and then licensed as text under CC BY-SA. This is probably the worst situation since there's no good license that goes with the code in this case. I'm curious if this is actually common though. I would be surprised if it turned out there are lots of people writing code that exists only on-wiki and isn't kept somewhere else. If this is common, I think it might be worth a different focus. It's actually hard to address this in the ToU because unlike the consistent CC BY-SA license there isn't one good software license for all of this. Off the top of my head, Apache, GPL, MIT, BSD all work fine for most things people would run on-wiki, for example. And different users might have good reasons for wanting slightly different software licenses. Something like a community policy that asks people on a project to license their original code under an acceptable OSI license would probably be better than a ToU update that tries to legally force this. If this is rare in the way I think it is, it might simply be worth flagging it to the code author when someone finds code with no specified license posted on-wiki, rather than anything more sweeping. -Jrogers (WMF) (talk) 16:47, 22 March 2023 (UTC)
Thank you for your detailed answer, @Jrogers (WMF):
I was mostly thinking on your third point and something along the lines of of wmf:Policy:Terms of Use/Phabricator, because JavaScript-coded gadgets (e.g. m:MediaWiki:Gadget-patrolRevisions.js) or Lua-code modules (e.g. Module:Arguments#L-1) exists mostly on-wiki only, and are code. Sometimes Python scripts (e.g. User:Ahechtbot/transclusioncount.py are made available as well). Posting MediaWiki extension's code on-wiki was also frequent (cf. mw:Category:Extensions which host their code in-wiki, although I think it's pretty rare these days; however mediawiki.org has lots of code fragments / snippets spread all over the site). Also, features that are now part of MediaWiki core or one of its extensions started as on-wiki Gadgets. How frequent this happens in terms of numbers is difficult to tell exactly. A quick estimate I got from the English-language Wikipedia via some SQL queries is:
  • 15,978 pages in the Module namespace.
  • And 414 pages whose title is MediaWiki:Gadget-[...] which host wiki-gadgets.
While not all of these pages may be hosting actual code, I'd say "on-wiki coding" it's not uncommon, but frequent.
Best regards, —MarcoAurelio (talk) 21:20, 22 March 2023 (UTC)

An edit

I made this edit so it points to the intended location of the wmf:Resolution:Licensing policy. The Resolution:Licensing policy of Meta points to m:Non-free content, which is related but not to the Board resolution itself which I think it's what you want readers to see. Posting here so Legal folks can review it. —MarcoAurelio (talk) 20:39, 11 March 2023 (UTC)

Transcript of #1 office hour

Hi all, we have uploaded the transcript of the first office hour. Thank you for those who attended, and we look forward to see you in the second one on April 4. Cheers, RamzyM (WMF) (talk) 00:27, 14 March 2023 (UTC)

Traduction en français / French translation

Point 8. Utilisation du terme "terminerons", phrase mal formulée. Ce sont les comptes qui sont terminés, non pas les utilisateurs !
« 8. Conformité DMCA (...) Conformément à la DMCA, nous terminerons, dans des circonstances appropriées, les utilisateurs et titulaires de comptes sur nos systèmes et réseaux qui ont enfreint de façon répétée les droits des autres. »
Proposition : « Conformément à la DMCA, nous terminerons, dans des circonstances appropriées, les comptes des utilisateurs et titulaires de comptes sur nos systèmes et réseaux, qui (...) »
Le verbe "terminer" est peu utilisé, "clore" ou "fermer" sont plus clairs, cependant le futur "nous clorons" est peu usité et peu compréhensible. Préférence donc pour "nous fermerons". MHM (talk) 09:57, 15 March 2023 (UTC)

Hah! Even by US copyright enforcement standards, that would indeed be somewhat drastic! Nosebagbear (talk) 10:07, 16 March 2023 (UTC)

Introduction, formulation en français

« Une partie de notre mission consiste à :
Encourager (...)
Diffuser (...)
Nous proposons Proposer des sites Web et une infrastructure technique pour vous aider à y parvenir. MHM (talk) 09:59, 15 March 2023 (UTC)

Gender-neutral language

Section 10 contains "his or her" and "he or she is". While we're making updates it would be good to change these to "their" and "they are" respectively, to be inclusive of individuals who identify as non-binary. the wub "?!" 12:32, 16 March 2023 (UTC)

We'll take a look, thanks! SSpalding (WMF) (talk) 01:24, 17 March 2023 (UTC)

Section 4: Concerns and Questions

In section 4, under the bolded section "Engaging in Disruptive and Illegal Misuse of Facilities", the section bullet reads as follows:

  • Engaging in automated uses of the site that are abusive or disruptive of the services, that violate acceptable usage policies where available, or and have not been approved by the Wikimedia community;

My questions about this specific bullet as as follows:

  1. There is a policy currently in development about OpenAI usage on the English Wikipedia. I'm not a lawyer. Does this bullet cover an as of yet not accepted policy?
  2. Does this sentence cover anything I wrote here? (I realize this is forward thinking and from the mindset of somebody with knowledge of automation.)

DarklitShadow (talk) 20:50, 16 March 2023 (UTC)

There's a strong implication here that it means violating an acceptable usage policy that is actually being enforced on wiki. One big picture thing to understand is that the terms of use is a tool to use to regulate people who have already violated project rules. So if it's not treated as operative by the community then the TOU wouldn't likely be implicated. To answer your questions directly, I'd imagine #1 would not be covered (in its current state). Definitely #2 wouldn't. SSpalding (WMF) (talk) 01:33, 17 March 2023 (UTC)

Change to CC BY-SA 4.0

It is a little surprising that on the explainer page you claim that "[a]s articles are updated, they will also become CC BY-SA 4.0." After all, in all likelihood, they won't (unless by "updated" you mean "replaced entirely with new content"). As Creative Commons points out, this has never been their understanding of their licenses. See creativecommonswiki:wiki/License Versions (stating that "[p]rior to the 4.0 versioning process, CC had not always been clear that the ShareAlike licenses stacked just as they stack for the BY and BY-NC licenses, and reasonable minds do differ on this point. CC believes, however, that this is the best reading of all of its licenses that permit adaptations prior to 4.0 and, now, has made that explicit in version 4.0."). Now, I myself beg to differ with them that "reasonable minds do differ on this point" since, in this jurist's view, any fair reading of the license compels this conclusion (cf CC BY-SA 3.0, § 4(a): "You may not sublicense the Work."), but be that as it may, one should certainly pay close attention to CC's legal position here.

Just some context for others reading along: The idea of "license stacking" is that if you license a work under license A and someone creates an adaptation under license B, then the adapter's license only covers the delta between the original work and the resulting work, meaning that any reuser of the adaptation has to comply with both A and B. Or in Creative Commons' words:
  • Creative Commons, Frequently Asked Questions: "If you make adaptations of material under a CC license [...], the original CC license always applies to the material you are adapting even once adapted [...] Recipients of the adaptation must comply with both the CC license on the original and your adapter’s license" (underlining added);
  • Creative Commons, 4.0/Treatment of adaptations: "When someone creates an adaptation of a CC-licensed work and licenses it, the license on the adaptation only covers the adapter’s contributions and does not extend to the original content. To the extent the original work remains distinguishable in the adaptation, the original is licensed to the downstream user directly from the original licensor. In other words, the person using the adapted work is the licensee under two separate licenses -- one from the adapter with respect to the new elements (i.e. the adaptation), and one from the original creator with respect to the original" (underlining added).
CC BY-SA 4.0 attempts to overcome these difficulties going forward through some (questionable) legal gymnastics by providing that "every recipient of Adapted Material from You automatically receives an offer from the Licensor to exercise the Licensed Rights in the Adapted Material under the conditions of the Adapter’s License You apply". CC BY-SA 3.0, however, contains no such provision.

Anyway, back to my initial point: With that being said, I think the intended change is misguided. Let's say the Creative Commons folks understand their license correctly. Then what you're about to do by dropping CC BY-SA 3.0 entirely from the Terms of Use and simply replacing it with CC BY-SA 4.0 is making it effectively harder to reuse Wikimedia content going forward. It will give rise to three categories of articles:

  1. CC BY-SA 3.0-only articles, which are all existing articles without a new (sufficiently original) addition made after the Terms of Use change;
  2. CC BY-SA 4.0-only articles, which are articles created from scratch after the Terms of Use change;
  3. CC BY-SA 3.0 articles with CC BY-SA 4.0 elements, which are articles created before the Terms of Use change with a (sufficiently original) addition made after the Terms of Use change.

For a very, very long time to come, a vast majority of Wikipedia articles and articles on other Wikimedia projects will fall into categories (1) and (3). In effect, therefore, the change makes the lives of reusers more, not less difficult. I shall note that, curiously, your new terms still require all content to be released under GFDL (effectively version 1.3), making GFDL the only fallback license that applies to old and new content alike. Given the known difficulties with having to obey GFDL as a reuser, this begs the question of why you've opted for a CC BY-SA 4.0/GFDL dual license rather than, say, a CC BY-SA 4.0/CC BY-SA 3.0 dual license, or a CC BY-SA 4.0/GFDL/CC BY-SA 3.0 triple license. I'm generally not partial to multi-licensing but if you accept GFDL as a historical debt in your mental accounting, at least having a practically useful "anchor license" like CC BY-SA 3.0 seems like a preferable option.

If it were for me, though, I probably wouldn't move to CC BY-SA 4.0 to begin with. I mean, if we'd have this discussion before the start of Wikipedia, I'd support using CC BY-SA 4.0, but the idea that you can just casually switch probably close to a hundred million pages written by unknown users under a roughly 3,000-word license agreement over to an entirely new roughly 3,000-word license agreement by just replacing a "3" with a "4" in the Terms of Use seems oddly simplistic. Just because the complexities associated with such a transition are not mentioned doesn't mean they aren't there. Even under your view that "as articles are updated, they will also become CC BY-SA 4.0" people would be stuck with CC-BY-SA 3.0 for a long time. So what would a recommended license notice look like after the adoption of the new Terms of Use? How do you explain the resulting licensing situation to reusers? Unfortunately, I don't see a discussion of any of that.

Best, — Pajz (talk) 22:56, 16 March 2023 (UTC)

Creative Commons has a very clear upgrade guidelines document discussing this. Quoting: CC has published guidance about what differs between versions. As a practical matter, compliance with a later version of a license is typically compliant with the earlier license. and For, example, reusing a 4.0-licensed translation of a 3.0-licensed essay would require complying with the conditions of both versions 3.0 and 4.0. This means attributing the original author as specified in v3 and attributing the translator as specified in v4. All Wikipedia needs to do is follow the guidelines of mark[ing] which license version applies to which content. CC has published best practices. Cheers! Freedom4U (talk) 09:59, 17 March 2023 (UTC)
Thanks for the link to the upgrade guidelines document, but I don't quite see what you're getting at since they make some of the exact same points that I raised above. It too pushes back against the claim made on the explainer page (and the Creative Commons 4.0 consultation FAQ), observing that "[i]n many cases, a work will have multiple rights at play because there is more than one contributor to a work. This often means there are multiple license versions applicable to a work. In those cases, reusers must comply with all relevant license versions when reusing the full work."
We can go through this in very practical terms, really. You write an article today. Tomorrow the Wikimedia Foundation changes the Terms of Use as proposed. Then I significantly expand your article. Next week, Joe finds the article and intends to use it on his website. Great. So what does Joe write in the notice accompanying the text? If he writes "This text is available under CC BY SA 3.0", that's not sufficient. CC BY-SA 4.0 § 3(a)(1)(A)(iii) (stating that "If You Share the Licensed Material (including in modified form), You must [...] retain the following if it is supplied by the Licensor with the Licensed Material: [...] a notice that refers to this Public License"). If he writes "This text is available under CC BY SA 4.0", that's not sufficient either because, under CC's (correct) view, he also needs to follow CC BY-SA 3.0 for his use of your original contribution, and then see, inter alia, CC BY-SA 3.0 § 4(b)(III) (stating that "You must keep intact all notices that refer to the Applicable License [scil. "this License"; see Legal code issues & errata] [...] with every copy of the Work as included in the Adaptation You Distribute or Publicly Perform"). So if Joe doesn't want to risk legal repercussions, he will have to make reference to both licenses now. But in order for Joe to know this in the first place, he would have to scour the version history of the article, whilst somehow knowing about the Terms of Use change and its effective date. This appears entirely impractical and that is why I wrote that I miss some discussion of how the Foundation envisages Joe's notice to look like.
In other respects as well, Joe's situation will become more difficult to navigate. For instance, he may have read that the CCPL4 license suit no longer requires him to provide the title of the work. That again is not helpful to him unless he knows -- and, again, if nothing but the Terms of Use is changed then he could only learn that by studying the version history and having positive knowledge of the Terms of Use change -- whether that actually helps to him because his article of interest is one of the few that is governed exclusively by the terms of CC BY-SA 4.0. However, as indicated, in all likelihood the article is not CC BY-SA 4.0-only, and so telling Joe he can simply follow CC BY-SA 4.0 would expose him to legal risk. The same applies to the other 'innovations' of the CCPL4 license suit.
So when you write that "All Wikipedia needs to do is follow the guidelines of mark[ing] which license version applies to which content," then yes, that sounds like a good plan, but how this is suppposed to work is beyond my imagination. If you take a hard look at the guidelines, their examples make abundantly clear why on project like this 'upgrading' is going to leave behind a chaotic patchwork of differently-licensed content. Creative Commons' only road to a clean upgrade with multiple contributors is having all prior contributors consent to the upgrade, and the practical example they give is from CC's own Terms of Use, which provide that users license their contributions under "a Creative Commons Attribution 4.0 International license or any later version of a Creative Commons Attribution International License". Well, there is no any-later clause in Wikimedia's Terms of Use. — Pajz (talk) 10:14, 18 March 2023 (UTC)
I echo your general sentiment that as one of the largest corpuses of CC content, this seems important to get right. A lot of this was debated in the 2017 consultation on Meta. The most relevant response is this legal note from WMF. To the extent your concerned with CC's own interpretation of our strategy, we've been in touch with the Creative Commons throughout this several year process. A lawyer from the Creative Commons organization was even available at the live community consultation three weeks ago to answer community questions about the strategy around this. Thanks for the feedback. SSpalding (WMF) (talk) 17:38, 21 March 2023 (UTC)
Thanks, SSpalding, I'm not quite sure what the "this" in your second sentence refers to but in case you are referring to the specific point that I've made, then I would just say that I did not find that aspect discussed as part of the consultation you reference; and the Wikimedia Foundation's conclusory statement in the legal note does not give any reason for their assumption. And, no, I'm not concerned about CC's interpretation of your strategy; I think I've made it clear that what I am concerned about is a legal issue that I fear will be caused by the adoption of your proposal. To recap, most of the argument is rooted in your implicit assumption that adapting CC BY-SA 3.0 content and licensing the adaptation under CC BY-SA 4.0 means that the resulting work can simply be used under CC BY-SA 4.0. I pointed out that Creative Commons makes clear across its website that this is not so. But my point is a legal one, really, and is not contingent on what CC thinks. It is also an observation made in the legal literature. See eg Melanie Bosshart's 350-page dissertation on the Creative Commons licensing system (Das Creative-Commons-Lizenzsystem, Diss., Zürich, Univ., 2013, Page Template:Different print/styles.css has no content.DOI:10.5167/uzh-164229doi:10.5167/uzh-164229) (stating that "the fact that the author of the original work retains rights to an adaptation makes Creative Commons licensing of adaptations complicated. It must be clear to the user that he or she must not only observe the licence of the adaptor but also that of the licensor of the original work", p 131; my translation) and Till Kreutzer's Practical Guide to Using Creative Commons Licences (available here on Meta), which Wikimedia Deutschland commissioned following the introduction of the CCPL4 licenses (stating that "[i]mportantly, any adaptation of a work still contains the original work. From a legal perspective, the adapter can only licence their modifications; unmodified parts of the work remain under the initial licence. Without a legal solution offered by the licence, the adapter cannot really 'relicence' the work as a whole. This might lead to the confusing situation in which the user of a repeatedly modified work has to obey multiple licences at the same time. The CCPL4 contains a new rule, which offers a simple solution for this problem [...]", p 34). So, no, this is not about whether Creative Commons is now endorsing your proposal or not. Anyway, I get the sense that your decision to drop CC BY-SA 3.0 entirely is a done deal at this point, so I'll just leave it at that. — Pajz (talk) 17:10, 22 March 2023 (UTC)
Just wanted to say that I am also confused and worried about these coming changes to CC BY-SA 4.0. In general, I'm unclear whether articles edited after the change will be (A) completely and entirely (re-)licensed under CC BY-SA 4.0 (as section 4(b) of CC BY-SA 3.0 seems or allow), OR (B) these articles will consist of a nightmarish "mix", in which some sentences are available under 3.0 and others under 4.0.
As an example, I have a project that uses data from dumps.wikimedia.org, the licensing of which seems to essentially defer to Wikipedia's general Terms of Use (basically this page). My project uses the top X most-used words on Wikipedia (which obviously draws from new and old article alike). I currently license it under CC BY-SA 3.0, following Wikipedia text's current Terms of Use. My question -- and I hope is common enough to be relevant to this discussion -- is: After this proposed change, will I be able to switch the license of my project to CC BY-SA 4.0, even though some articles will remain unchanged since the change in Terms of Use? (I will say that I don't think "dual-licensing" under 3.0 or 4.0 makes legal sense, because, in my understanding, that is effectively giving users and would-be adapters a choice between the two licenses.) Survive morse mocker (talk) 22:08, 14 April 2023 (UTC)

Wikinews, marketing company mediation

Hi! Thanks for thoroughly detailing these updates. I noticed that in the Licensing of Content section, the wording about Wikinews and its sole (no GFDL dual-licensing) CC BY 2.5 license was deleted; does this mean that Wikinews will transition to a CC BY SA 4.0 and GFDL license? I also see that there is no mention of Wikidata's CC0 licensing in the section - is there any specific reasoning behind this?

Additionally, the term Marketing Company Mediations is used often in the Terms; I can't seem to find online any documentation of what it means. Could you please clarify its meaning?

Thanks! EpicPupper (talk) 22:17, 17 March 2023 (UTC)

#Marketing Company Mediation above might be worth reading irt. your second question — TheresNoTime (talk • they/them) 00:29, 18 March 2023 (UTC)
Thanks for the link, @TheresNoTime! @SSpalding (WMF), would these Marketing Company Mediations be only applicable to marketing companies (as per their name)? While paid editing agencies make up a significant portion of the UPE landscape, individual companies also problematically instruct their employees to edit articles without disclosure, sometimes continuing after attempted intervention. Would these companies be subject to the med-arb process? Best, EpicPupper (talk) 21:32, 19 March 2023 (UTC)
There's a simple answer and a more complex answer. The more complex answer is probably best explained in all of the other conversations so far on this discussion page.
The simple answer is that we are enforcing against "marketing companies" because they work systematically and continuously, they have the fewest arguments that they were acting in good faith, they have the most to lose if they don't comply, they are most likely to be interested in following the rule of law (and listening to court orders), etc.
That said, th
hese companies often end up hiring "individuals". Some of these Upwork profiles, for example, get clients from larger companies that collect clients but then sub-contract the work to others. If we limit the ability for the companies to collect clients, there will likely be less work for these individuals to be contracted to In short, there's a wide range of gray area activity that is best left up to the community to use it's discretion about when it comes to day to day individual edits on the project. But this is for wide spread systematic violations there's obvious bad faith mixed with commercial incentives.do. SSpalding (WMF) (talk) 21:57, 19 March 2023 (UTC)
The default move to CC BY SA 4.0 would not affect any project that expressly chooses to have a different licensing scheme. Said differently, in the same way that Wikimedia Commons already allows CC BY SA 4.0 contributions, Wikinews and Wikidata's communities will keep their bespoke licensing schemes unless the communities choose otherwise. SSpalding (WMF) (talk) 03:33, 18 March 2023 (UTC)

Interconnected Wikipedia articles debunking biases from different ‘provincialist’ views

Wikipedia, sooner or later, will have to counter & debunk myths, legends, misinformation, propaganda, disinformation, and forms of subtle narrative favouritism. There are articles that do not have a dedicated English page that are “hidden” to that specific language or a specific limited set of languages. Then, the issue of the content itself. Eventually, to avoid incomplete, misrepresented or lack thereof of article content, there should be a measurement attentioning the reader that “This article is shorter/longer/similar/lack of references, bibliography, annexes, and other sources.

This could provoke reactions from governments, establishments, even dedicated state-sponsored state entities to maintain a certain narrative, and readers will not check the other versions of the article(s) in different languages to analyse & deduct what's missing/omitted, what's alterated, what's censored, what's differently constructed (misinformation, disinformation...). As a light mainstream example, when discussing history, “presentism” can become an obstacle in itself mixed with provincialist flavour (eg. “The migration populations arrived in Europe starting with abcd CE Century” versus “The invasion of wxyz populations occured at the beginning of the abcd CE Century”). None of the statements are wrong, however, there are provincialist flavours that cannot be conciliated in and of themselves. Proven to be the lingua franca, the English version will have to include provincialist views as a form of a "complete" version instead of cherry-picking based upon the reader/editor/writer's biases or entrenched convincements that are, themselves, backed by institutions' chosen narrative(s). Trexerman (talk) 01:03, 20 March 2023 (UTC)

Siezen oder Duzen? / Formal or informal? (on DE translation)

In der de-Wiki duzt man sich generell. In diesen Nutzungsbedingungen wird aber gesiezt. Ich finde, dass wir weiterhin beim Du bleiben sollten. Abubiju (talk) 11:40, 20 March 2023 (UTC) At De-Wiki you are generally on first name basis. In these Terms of Use, however, we are on formal terms. I think that we should continue to stick to the "Du" (=informal). Abubiju (talk) 11:40, 20 March 2023 (UTC), translated by DBarthel (WMF) (talk) 13:14, 23 March 2023 (UTC)

Déjà moi je ne connais pas parler ta langue Mathis4545 (talk) 14:26, 21 March 2023 (UTC)
Tatsächlich existiert für die Sie-Form ein spezieller Sprachcode (de-formal). Beim gewöhnlichen de wird daher normalerweise die Du-Form verwendet. ~~~~
User:1234qwer1234qwer4 (talk) 09:26, 23 March 2023 (UTC)In fact, there is a special language code (de-formal) for German formal terms. It is common in de-wikis to use the informal terms. ~~
User:1234qwer1234qwer4 (talk) 09:26, 23 March 2023 (UTC), translated by DBarthel (WMF) (talk) 13:14, 23 March 2023 (UTC)

22 March 2023: Update notice

Hi all! We have updated the Proposed Update (comparison of the versions) and posted an explanation about it, based on the feedback from this talk page and from the first office hour. Please keep the feedback coming on the current version of the Proposed Update below this message. Thank you very much for your participation. Best, RamzyM (WMF) (talk) 11:16, 22 March 2023 (UTC)

It seems some formatting of del/ins is lost. Wargo (talk) 20:29, 22 March 2023 (UTC)

Paragraph about history page

Please, describe this text: (since each page has a history page that lists all contributors, authors and editors) in a separate paragraph with more information about the role of each edit in attribution. For example, ToU must directly say if an edit summary must provide attribution for importing text (or attribution may be in the edit itself). I propose addition like this:
If a contributor makes an edit that imports text, he must provide attribution within the edit, otherwise attribution must be provided in another edit that refers to the edit without attribution. D6194c-1cc (talk) 06:13, 26 March 2023 (UTC)

Thanks for this suggestion, we'll take a look at it. I will say I'm not sure when looking at this suggestion if it would be good to specify this level of detail in the ToU. It may be that different projects have different preferences on how to do this. For example, if text is imported with attribution in a footnote, it's not clear legally that it would also require attribution in an edit summary. If we put something in the ToU, it's requiring a lot of uniformity across multiple projects and languages, so I want to look at that carefully. -Jrogers (WMF) (talk) 19:17, 30 March 2023 (UTC)
If it will be in ToU, it will put the end of misunderstanding about history page role. Many think that they can import text without any attribution in the edit, but when someone adds a piece of text, he licenses it under CC BY-SA. So each re-use of this text must be supplied with attribution and information about modification (at least whether it was modified). If the imported text has no attribution in the history page, then ToU and license violation occurs. History page helps to determine whether the text was modified or not. D6194c-1cc (talk) 12:00, 1 April 2023 (UTC)
Probably, clarification like within the edit (whether in the edit itself or in the edit summary) might solve the problem, but it has some legal concerns. Attribution in the footnote might be the right way for importing from external sources, which doesn't comply with Wikipedia ToU. Also, attribution in the edit summary helps to check page history for imported text without the need to check every edit contents. CC BY-SA 4.0 requires that attribution will be provided "in any reasonable manner", but checking every edit content probably will not be reasonable in case of history page (that still lists all contributors). So mandatory edit summary attribution seems like the best choice to satisfy both the ToU and license. D6194c-1cc (talk) 12:23, 1 April 2023 (UTC)
Although, the official reply by Samwalton9 (WMF) from the Wikipedia Legal team about the role of edit history, as I understand: ru:special:diff/129686380. @Jrogers (WMF): Can you comment this reply in context of ToU and the "(since each page has a history page that lists all contributors, authors and editors)" claim? D6194c-1cc (talk) 07:42, 7 April 2023 (UTC)

german version: gender neutrallly wording

in general the using of gender charaktes, "*" or ":" or "I" is ok, but for german wiki there is a dicision available: https://de.wikipedia.org/wiki/Wikipedia:Generisches_Maskulinum JanEhlebrecht (talk) 09:23, 28 March 2023 (UTC)

@JanEhlebrecht - danke für den Hinweis. Da die von dir verlinkte Entscheidung sich ja auf den Artikelnamensraum bezieht, sollte das kein Problem darstellen. Ganz liebe Grüße, Thanks for the hint, Jan. Since the decision you linked refers to the article namespace, that shouldn't be a problem. Best regards, DBarthel (WMF) (talk) 10:03, 28 March 2023 (UTC)

Other countries' laws

We did not adjust the language in the overview indicating that laws may apply where “you or the subject you’re writing about are located).” As there was feedback about this, we want to explain why we have left it. We added this language to help warn editors to be careful about external risks, as we have seen several legal cases in the last decade that used the law of the article subject, rather than US law or the location of the editor. Unfortunately, this trend is likely to remain true for the foreseeable future whether we put it in the terms of use or not. For clarity, this language in the overview does NOT refer to Foundation enforcement actions. It’s only there to warn editors to be cautious when writing about living persons or organizations who might initiate a lawsuit.

Could we please clarify that violating the laws of a country other than the US is not a ToU violation, and state that the Foundation will resist efforts to prosecute such accusations? 2601:647:5701:39B0:3EA4:876B:A3AC:D588 09:53, 28 March 2023 (UTC)

Third party resources

I just wanted to reiterate, that i still object to the inclusion of a policy not yet written. I do not believe that there can be meaningful consultation on a section that is essentially TBD. Additionally, while the change from uploading to improper use is an improvement, I still think its misleading since the policy is allegedly not about using third party resources, but about transmitting personal data to third parties without consent (but hard to say for sure since the policy doesn't exist). Bawolff (talk) 17:14, 29 March 2023 (UTC)

Duly noted, I think this is a fair critique. We may end up taking this section out by the end, I'll plan to flag this to the team working on it. -Jrogers (WMF) (talk) 21:11, 29 March 2023 (UTC)

Seconded. Nemo 10:37, 31 March 2023 (UTC)

"Intellectual Property Rights"

No such thing exists. The addition of this term adds no value compared to the already existing enumeration of the specific relevant regimes. Nemo 10:01, 31 March 2023 (UTC)

English Wikipedia specific content

The proposed change would introduce a link to w:en:Wikipedia:Plain and simple conflict of interest guide. I have no idea what this means. The terms need to apply to all Wikimedia projects and must not include anything that is specific to the English Wikipedia. It's especially inappropriate to incorporate some English Wikipedia policy or guideline by reference, let alone an essay. Nemo 10:06, 31 March 2023 (UTC)

I can't agree with you more. Except for this link, I think mediations are at least two benefits, such as require reimburse and talk about cooperate to improve wikis in the future. Long5563 (talk) 19:37, 6 April 2023 (UTC)

Don't use "click here" as link text

Per https://www.w3.org/QA/Tips/noClickHere, the proposed text contains various inappropriate uses of "here" as link text. For example "Our contact details are available here" should just be "Our contact details are available" or "See our contact details". Nemo 10:10, 31 March 2023 (UTC)

Thanks for pointing this out. In our explanation, we explained we did a first pass to try to make links accessible to screen readers. This is one that it looks like we missed. I've made a note of it to fix in the final version. SSpalding (WMF) (talk) 17:50, 3 April 2023 (UTC)

Repeat copyright infringement

The proposed update would remove "repeat copyright infringement" and replace it with "repeat posting of unlawful material under applicable law in line with human rights principles". What we consider copyright infringement is often not unlawful material. For example a "copyvio" on Commons can easily be fair use in the USA. Does this mean that the WMF is no longer interested in banning users who engage in such activities?

In my opinion the previous text was better. What we care about the most is preserving the availability of the projects under their free license. So respecting the license is the primary objective of the terms of use. Mentioning copyright infringement explicitly makes this clear. On the other hand, conflating all sorts of laws provides no value (what's illegal is going to be illegal whether it's in the terms or not) and adds confusion, such as the need of specifying what human rights principles might or might not apply (speaking of which, one can easily argue that the existing USA/EU copyright regime doesn't respect human rights or the First Amendment; that doesn't mean we're going to give up on our licenses in the meanwhile though). Nemo 10:22, 31 March 2023 (UTC)

This proposed update revisits once again the issues of "Paid contributions without disclosure". The introduction of this section has failed to serve any useful purpose and has only multiplied uncertainties on the projects. It's time to drop the stick and back slowly away from the horse carcass. Just remove the entire section and adopt a different approach. This is a complex situation and it will not be solved by lawyers. Nemo 10:33, 31 March 2023 (UTC)

what different approach would you like? --ThurnerRupert (talk) 17:09, 31 March 2023 (UTC)
One that works. :) I don't have a preferred solution. The first step would be to define the goals. The next would be to task an independent third-party with researching what methods have been working better (people involved in marketing on Wikipedia will need a promise of confidentiality). Then the well-functioning methods can be promoted more, as needed. (Maybe it's just a matter of education?)
I hope it's clear to everyone that lawyers and private legal action are not the only way to solve problems in the world. As I see it, if people manipulate the Wikimedia projects to deceptively alter content to their private benefit, the damaged party is the public, not the WMF. So ideally there would be ways for the public to defend itself. For example, if we're talking about the USA, political campaigns abusing Wikipedia could be subject to FEC intervention, while businesses are subject to the FTC (which has already established rules about "influencers"); in the UK, one can easily report deceptive advertising to ASA/Ofcom. Nemo 16:46, 20 April 2023 (UTC)

Illegal Content

"Lawful Behavior — You do not violate copyright, post illegal content or violate other applicable laws that follow human rights principles. "

From my point of view, "illegal content" depends on the laws of the countries, and may change over time, so I think it is a point to consider.

--Pilokyoma (talk) 16:22, 31 March 2023 (UTC)

Although illegal content is vague, but I think the update revision clear than the current version. Long5563 (talk) 11:59, 1 April 2023 (UTC)

Mnmh... Section 10

Let me start by thanking you the hard work and skill on the changes and improvements made here and for the stewardship of the project generally, a stewardship that is now 20 years on and has seen the project grow to become a true ornament to humanity.

OK, moving on, I see that Section 10 (Management of Websites) expands quite a bit on on the Foundations powers to kick out people. That might (or might not) be fine, but I note the following passages (emphasis added):

  • "If you contact the Foundation with a problem, we will typically explore whether and how existing community-led mechanisms can investigate and, where appropriate, resolve it".
  • "In an unusual case, the need may arise... to address an especially problematic user or especially problematic content because of significant Project disturbance"
  • "Those Foundation moderation activities may be informed or performed by software (such as traffic flood (“Denial of Service”) protection). In those cases human review is normally available, upon request.
  • "Other information about routes of appeal may also be explained to you at the time"

But you know, real contracts don't have this. There isn't "The homeowner will normally be liable for damage caused by raccoons, unless in his judgement it's an unusual situation." Maybe contracts do sometimes have "in most instances" and so on,IDK, IANAL. But if so, 'these contracts assume access to adjudication by a judge if it comes to that.

"If you contact the Foundation with a problem, we will typically explore whether and how existing community-led mechanisms can investigate and, where appropriate, resolve it; but we don't have to -- if our judgement the situation is atypical in some way, or anyway we can say so, we will do whatever we want."

I mean, that's what it says (effectively), right? Ask the lawyers.

I get that most people reading this are going to be like "Oh, come on, you know this was written in good faith to provide for the application of common sense in some unforseen cicumstance that may arise, because after all a the law is a blunt instrument and it would be dumb to have a rule that forces the Foundation to do something that's silly". I know that committee members reading this will probably be like "Well, this is insulting, if not actually trollery, we would never use these clauses to do anything wrong, why is this person assuming otherwise?

But ask a lawyer. Anything in a contract that is expected to be generally understood by plain horse sense because, you know, we're all friends here... if it isn't written down, it isn't worth the paper it's not written on. If a lawyer says otherwise, she probably shouldn't be lawyering.

This is not theoretical either. The Foundation already disappears people and doesn't say why. Fram and there was another guy later IIRC. Looks like this stuff just locks down that right. Doesn't it.

Willing to have my mind changed, but at this point, thumbs down. Herostratus (talk) 18:05, 31 March 2023 (UTC)

Thanks for the feedback. The nuances you bolded are there to represent the reality: that the TOU is not some blunt instrument that would be wielded irrationally, but it's also impossible to describe every exception to the general rule in one sentence. Taking two of your examples:
  1. "Typically" might refer to a response to CSAM material which didn't get caught by community process but was reported to us that we have legal obligation to take down immediately upon notice. A more mundane example might related to a DMCA notice which we also have an obligation as a host to takedown if the notice is made in good faith and checks all of the requirements boxes. A forward-looking example might be "any future law that the projects are governed by that require action from the host" passed in the next few years. The last significant update of the terms was 10 years ago. At that rate, it's almost certain that there will be a different set of intermediary laws governing hosts that we will have to account for. Thankfully, the projects get very few notices of these kinds as compared to other platforms because of the responsive community, but these are still exceptions to the rule and there will always be because of laws like this.
  2. "Normally" might relate to security concerns. If a Bad Actor is trying to DDOS the projects and they are frustrated by project security, it would be very unwise to provide them -- upon request -- with logs detailing how security responded so they can do a more successful DDOS in the future.
I don't want to discount the Fram situation. Although I was not at the Foundation at the time, I've learned that this was a major piece of contention that soured many community members' relationship with the Foundation. That said, from everything that I do day-to-day when started 2 years ago, there seems to be an express effort to avoid making those types of decisions today. I realize that many might want to debate that, but that's my limited, personal experience.
In sum, the nuances you highlighted really do represent good faith ways of making general statements that reflect reality while also (1) avoiding dozens of context-specific exceptions that might be out of date or unresponsive to the legal climate in another 10 years, (2) avoiding having to change the TOU every quarter to add new trivial exceptions or erase out-of-date ones. SSpalding (WMF) (talk) 18:39, 3 April 2023 (UTC)

Lengthy issue for paid disclosure

I suggest the following alternative:

  • In addition, if you make a public posting off on the Wikimedia Projects advertising editing services on edit Wikipedia in exchange for compensation of any kind, you must disclose all Wikipedia accounts you have used or will use for this edit service in the public posting on the third-party service.

Disclosure paid contribution is include all edit, not only advertising even if they are suspicious. Undisclosed paid editor may remove objectionable content or post legal threat to defend their interest.

  • Applicable law, or Project specific policies and Foundation’s policies and guidelines, such as those addressing conflicts of interest, may further limit paid contributions or require more detailed disclosure. For example, infringing Wikimedia Trademarks (defined hereinafter at Section 6) to advertise paid editing services, removing disclosures on previously adequately disclosed paid edits, or logged-out paid editing in a manner that makes adequate disclosure a practical impossibility violate this section.

Many examples can write in Terms of Use, but it should only keep summary and put examples to appropriate policies or guidelines. I don't know about logged-out paid editing. Long5563 (talk) 19:11, 1 April 2023 (UTC)

@Long5563 (how about second chance )/ 174.215.149.238 22:55, 1 April 2023 (UTC)
The section of “paid contributions without disclosure” have not mentioned something about second chance, could you tell me more detail? Long5563 (talk) 04:10, 2 April 2023 (UTC)
Change a little bit:
In addition, if you make a public posting off on the Wikimedia Projects advertising editing services on edit Wikipedia in exchange for compensation of any kind, you must disclose all Wikipedia accounts you have used or will use for this paid editing service in the public posting on the third-party service.
Thanks for your feedback.
  1. The text already prohibits "making edits" on Wikipedia in exchange for compensation elsewhere. This is the biggest problem. The sentences you are pointing out actually adding a different, additional requirement. This part prohibits "advertising UPE services." It looks like there might be a typo that needs fixing. Thanks for pointing this out.
  2. These examples are actually in response to earlier feedback from the community at that link and elsewhere asking us to be specific about those things even when they are already implied. They were included because a handful of people asked for and +1'd the clarification. 3. As for "second chances," I don't know if you're talking about a specific concept or existing rule. If you're asking about giving people who conduct systematic UPE campaigns second changes generally, here's my take. The Foundation will only be enforcing this section is egregious circumstances where the company is systematically and intentionally being a bad actor that is NOTHERE to help further free knowledge. In the situations where it gets to the point where a company receives a formal notice of mediation, they have likely proven to be already proven to be a problem to the community.
SSpalding (WMF) (talk) 17:48, 3 April 2023 (UTC)
1. The section of “Harassing and Abusing Others” has been forbidding spam. The term of “advertising” maybe misleading to assume that allow paid editing if they do not violate NPOV or Notability. 2. Local policies or guidelines have different clause, therefore examples are not applicable for some wikis. @User:SSpalding_(WMF): Could you tell me what is logged-out paid editing? (Thanks for your attention to this matter, you can ignore the question if you unwilling. ) Long5563 (talk) 19:37, 6 April 2023 (UTC)
EN:WP:G11, indicate the speedy deletion criteria of unambiguous advertising or promotion. Long5563 (talk) 16:07, 17 April 2023 (UTC)

การรองรับการติดต่อเรียนรู้แก้ไขปรับเปลี่ยนสู่"วิกิพีเดีย"2023

เรียนท่านผุ้อำนวยการ ผู้บริหารและกลุ่มเจ้าหน้าที่กรรมการหน่วยงานต่างๆขององค์กรและมูลนิธิ"วิกิพีเดีย"ทั่วทุกมุมโลก เรื่อง เริ่มการอำนวยการอัปเดตข้อกำหนดการใช้งานและประกาศอัปเดตสำหรับการใช้งาน"วิกิพีเดีย"ว่ามีการปรับเปลี่ยนเรื่องใดๆบ้างที่สำคัญและจำเป็นต้องปฎิบัติตาม"กฎ ระเบียบ บัญญัติ"ที่ทาง"วิกิพีเดียได้ปรับเปลี่ยนและสร้างขึ้นเพื่อรองรับกับวิวัฒนาการใหม่ๆทางด้านระบบ"ปัญญาประดิษฐิ์"(AI)นำมาประยุกต์ใช้เพื่อให้ทันสมัยกับความเจริญก้าวหน้าของระบบและการปฎิบัติงานและการประมวลผลในรูปแบบต่างๆจากการบันทึกข้อมูลให้ได้คำตอบที่ถูกต้องและใกล้เคียงกับประวัติศาสตร์และสิ่งที่เกิดขึ้นในเรื่องราวหรือการพิสูทธิ์ทางด้านเทคโนโลยีและวิทยาศาสตร์ที่เกิดขึ้นจริงครับ

นาย อนุรักษ์ ศรีจันทรา อนุรักษ์ ศรีจันทรา90 (talk) 11:16, 4 April 2023 (UTC)

Google Translate-provided translation:
Support contact, learn, edit, change to "Wikipedia" 2023
Dear Director Administrators and groups of officers, directors, departments of "Wikipedia" organizations and foundations from all over the world on the commencement of providing updates to the Terms of Use and posting updates for the use of "Wikipedia". That there are some important changes and it is necessary to follow the "rules, regulations" that the way "Wikipedia has modified and created to support the new evolution of the system" artificial intelligence "(AI) is applied to keep up to date with the progress of the system and operations and processing in various forms from recording data to get accurate answers and close to history and What happens in the story or the proof of technology and science that actually happened?
Mr. Anurak Srichandra Anurak Srichandra90 (talk) 11:16, 4 April 2023 (UTC)

Alternative paid contribution disclosure policy

The ToU proposal says "A Wikimedia Project community may adopt an alternative paid contribution disclosure policy which shall preempt this section". So in spanish Wikipedia there is this policy about paid contributions procedure. This spanish policy says nothing about the new statement "if you make a public posting off the Wikimedia Projects advertising editing services on Wikipedia in exchange for compensation of any kind, you must disclose all Wikipedia accounts you have used or will use for this service in the public posting on the third-party service." Question: As long as they dont update the spanish policy, in spanish Wikipedia nobody needs to follow that statement from ToU, right? Cartago90 (talk) 10:29, 5 April 2023 (UTC)

We discussed projects with different UPE rules more in previous threads on the topic (the 20 February version). The summary is that we want to preserve the autonomy of projects that have intentionally chosen to adopt different rules (e.g. Wikimedia Commons) while making these TOU rules the default for projects that don't have specifically adopted rules. This has always been true even before the update.
There's a lot of nuanced discussion in those previous threads, but to answer your question at least somewhat directly: if a community has no problem with what a company is doing in that community related to undisclosed paid editing, then it'd be extremely unlikely that that the Foundation would have any interest in enforcing this section against that company. These updates are meant to correct systematic problems on wikis, not to create problems out of new non-issues.
So the most meaningful open question is: what does the Spanish Wikipedia community think? Should people on Spanish Wikipedia follow the statement or not? SSpalding (WMF) (talk) 18:35, 6 April 2023 (UTC)
I have opened a discussion in the Spanish Wikipedia about this: [10]. I highly doubt the Spanish Wikipedia community intends to make the paid editing requirements in Spanish Wikipedia more lax than the Terms of Use. MarioGom (talk) 09:09, 16 April 2023 (UTC)

Section 4

I want to ask, what is defined as "vandalism" as written in section 4? RXerself (talk) 15:37, 5 April 2023 (UTC)

Any action consciously and deliberately done to harm the project. Long5563 (talk) 19:37, 6 April 2023 (UTC)
Is that your definition or the definition mentioned somewhere in the Terms of use? RXerself (talk) 05:20, 20 May 2023 (UTC)
Vandalism is not defined expressly in the terms of use. What is and isn't vandalism is mostly defined by project specific policies: https://en.m.wikipedia.org/wiki/Wikipedia:Vandalism . What is acceptable on one wiki may be unacceptable on another. Should an action be of no concern to the community it affects, it'd be unlikely to merit reaction at the terms of use enforcement level. I hope that answers the question about the definition. SSpalding (WMF) (talk) 06:50, 21 May 2023 (UTC)

Section 7 Licensing: use and attribution of LLM and other AI content

I suspect that the legal department may already be having discussions about how to update section 7 Licensing (esp. b) Attribution, and c) Importing text, in order to provide guidance on what, if anything, must be attributed for non-human, LLM or other AI-generated content, such as content created by w:Chat GPT. This is already being discussed intesively at Wikipedia in dozens of forums, with the recognition that many articles have already been created or updated by users adding content created by Chat GPT; here's one recent discussion at ANI about a brand new editor who rewrote the leads on eight biographies in 28 minutes, and here's a proposal to create a policy concerning LLM-generated content.

In my opinion, an update to section 7 is urgently needed. As this discussion period for updating the ToU will continue until late April, and I imagine it will take at least some months to digest it all, and come up with a new ToU, imho that is way too long to wait at Wikipedia, to figure out how to deal with this content, which is sprouting all over. I would like to get some feedback in the meantime from WMF Legal in advance of the final ToU release to these questions:

  • Are you currently discussing how to deal with Chat GPT and other LLM content with respect to ToU section 7?[1]
  • Do you have a timeframe for updating the ToU, and any chance section 7 could come first, or sooner?
  • If the expression "creative content" is key here, will you be addressing that, wrt LLM/AI models? (§&bnsp;308.2, et al.)
  • Can you offer at least some interim guidance now, or soon, even if not set in stone, as to how you will be viewing LLM-generated content, with respect to attribution, and content importation? This is urgently needed.

I realize that U.S. law on Copyrightable Authorship restricts this to human authorship (§&bnsp;306), which implies that the floodgates are open (currently, given current ToU that discusses "creative content"), but if I understand correctly, there's nothing saying we can't be *more* restrictive than whatever governing U.S. law happens to be, hence I assume WMF may still have something to say about this, and soon, I hope. (U.S. law will eventually have to address this as well, but hopefully WMF is a lot more nimble than they are.)

This question was provoked by this subsection of a discussion at the Talk page of the proposed Wikipedia policy on LLM (w:WP:LLM). Thanks, Mathglot (talk) 04:26, 10 April 2023 (UTC)

Adding courtesy links:
m:Wikilegal/Copyright Analysis of ChatGPT
Mathglot (talk) 04:52, 10 April 2023 (UTC)
Thanks for your feedback. You're right that essentially every department at the Foundation has been involved in figuring out how to minimize the negative effects the commercialization of LLMs will have on the projects (while encouraging the positive ones).
For anyone who isn't aware of all of the different ways LLMs may affect the projects, we recently hosted a community call with extensive notes outlining opportunities and challenges. See also, the many conversations on project village pumps and community policy-making.
Although the international legal landscape around LLMs is changing almost weekly, that wikilegal article ( m:Wikilegal/Copyright Analysis of ChatGPT ) represents some, but not all, of the most durable principles related to LLMs and copyright law in the United States where the Foundation is headquartered.
Currently, the projects seem to be doing a good job creating policies, draft policies, and norms around this technology. See for example, this discussion on Wikimedia Commons. Intervening with something written into the terms of use, at this stage, might harm more than it would help. For example, there are both legal and important philosophical/ethical pieces to the question of how useful generative text and generative art is to a project. The philosophical questions are best answered by a consensus of project contributors who understand local norms and context. For example, LLM's present very different issues on English Wikipedia compared to a less-common language (where the LLM doesn't work very well because it wasn't trained on it) with a smaller editor base (who might be concerned with being overwhelmed with moderation tasks).
All that said, you will likely be seeing many more specific public statements, on Diff and elsewhere regarding various aspects of LLMs and how they affect the projects, the open licensing ecosystem, etc. Legal has been involved with helping craft some of these. SSpalding (WMF) (talk) 21:00, 11 April 2023 (UTC)
To treat this as a copyright issue would be the wrong way. Editing articles with the use of AI — editing by non-humans in general — should simply be prohibited. Just put this in the Terms of Use, point. How to identify AI-texts is step two; may be, there can be developed a tool for it. --Wickey (talk) 14:02, 12 April 2023 (UTC)
Mathglot's initial post included an example of what we currently believe best addresses generative text bots: "This question was provoked by this subsection of a discussion at the Talk page of the proposed Wikipedia policy on LLM (w:WP:LLM)." Projects are currently devising their own rules and norms beyond copyright law. Editing articles with the assistance of generative text algorithms doesn't implicate any "novel" legal principles at issue that are relevant to the projects, so the TOU may not be the best place for including rules governing them.
The needs of English Wikipedia are very different than the needs of Wikimedia Commons so local rules should be created rather than creating a project-wide rule. For example, it's definitely correct to say that generative text algorithms may create new vectors for mis/disinformation. But users intentionally introducing disinformation is already prohibited by terms of use. Existing local project rules (e.g. accurate citations, notability requirements, etc.) also address disinformation, and those can also be enforced via the TOU.
There's a very long history of human editing assisted by non-human tools: for example, bots that spot typos and do other rote functionary work. To create a complete "ban" on scripts that help contributors contribute could have an overall negative effect.
Finally, yes there is a need to figure out technical ways of identifying generated text since that seems very important to the future information ecosystem. Wikimedia projects should hopefully be meaningful part of that conversation as researchers try to address that question. SSpalding (WMF) (talk) 16:59, 14 April 2023 (UTC)

The US Copyright Office has been looking at this as well, including the human authorship issue, and had this to say recently:

{{{1}}}

Mathglot (talk) 07:52, 19 April 2023 (UTC)

Notes

References

  1. ↑ Sounds like the answer to this is "yes": m:Wikilegal/Copyright Analysis of ChatGPT.

Thank You

For work !! Zemant (talk) 17:44, 10 April 2023 (UTC)

Thanks for the support! SSpalding (WMF) (talk) 16:24, 14 April 2023 (UTC)

Section 7 - citing Wikipedia

"When you re-use or re-distribute a text page developed by the Wikimedia community, you agree to attribute the authors in any of the following fashions:

1. Through hyperlink (where possible) or URL to the article 2. Through hyperlink (where possible) or URL to an alternative, stable online copy or 3. Through a list of all authors (but please note that any list of authors may be filtered to exclude very small or irrelevant contributions)."

Now this is interesting. At the moment e.g. journalists, students and bloggers say "as it's written on a Wikipedia article: ...". So in future this wouldn't be allowed, a name of the article is not enough but they would need to use the url of the article instead? And do we help them to get a list of authors of a certain chapter of a Wikipedia article? Jjanhone (talk) 18:33, 11 April 2023 (UTC)

Well, according to the ToU it wasn't and it won't be guaranteed correct way to attribute information from Wikipedia without hyperlink or authors list. The attribution methods in the proposal didn't changed, except for the added "contributors" to the explanation. Without attribution any of article authors can complain about violation of the license and his copyright (since Wikipedia doesn't own copyrights on added texts). D6194c-1cc (talk) 20:25, 11 April 2023 (UTC)
Although, attribution just by URL or just by name is vulnerable to article renaming/deletion. Probably, the most correct way to attribute text form Wikipedia is to give a permanent link to the actual article revision. D6194c-1cc (talk) 20:32, 11 April 2023 (UTC)
The last is obviously true: a permanent link. --Wickey (talk) 13:39, 12 April 2023 (UTC)
After I posted this I realized that this wasn't anything new. But as I see that this is not happening at all at the moment but people are just using the article name "like they say in Wikipedia article X...", the rule is not working. I'd like to see a button on Wikipedia, next to the article title or even next to each chapter with a question mark: "Want to see who wrote this? Click here for more". Many readers never visit history article.--Jjanhone (talk) 06:48, 13 April 2023 (UTC)
Well, history page is not only for the attribution. Suggested CC BY SA 4.0 is much less risky in context of attribution practices. CC BY-SA 4.0 says: {{...indicate if You modified the Licensed Material and retain an indication of any previous modifications; and...}}. So providing a link to the article we provide information about authors and about all previous modifications of the work. But this definition of attribution conflicts with the 7(b)(iii) section of the ToU, because authors list misses information about previous work modifications. But probably ToU is superior in this case. CC BY SA 3.0 didn't explicitly said that information about modification is part of attribution; as I understand usually it means just authors. D6194c-1cc (talk) 11:03, 13 April 2023 (UTC)
Also, the idea about authors list for a specified part of text is great. It requires some technical work to add necessary functionality to the engine. When copying some text from Wikipedia that can be separated from other text in the page history, this text probably can be used independently, because each of the authors contributed in this text published it under CC BY-SA, so each of the authors is an exclusive author of his piece of text unless someone else will contribute.
The only problem is in legal context about authors that add changes among multiple parts of the article. The question is whether a piece of text added by author A can be used separately if another author B made style changes across the entire article in a single edit. I hope that somebody from Wikipedia Legal team will add a comment about this issue. D6194c-1cc (talk) 12:26, 13 April 2023 (UTC)
Since this isn't "new" text (it's already in the existing 2016) version of the terms of use, can you be specific about what in this discussion you believe are the most pertinent things someone from the Foundation Legal department can comment on here. Happy to comment, but I just want to make sure I know what's the core issue that needs to be addressed / clarified.
I'll try to address this, "whether a piece of text added by author A can be used separately if another author B made style changes across the entire article in a single edit," as directly as possible even though clarification would be appreciated:
1. Although I wasn't around for the 2010 drafting of the 7(b)(3), "Through a list of all authors..." I imagine this was for edge cases: in situations where URLs or other attribution methods weren't feasible. Anecdotally, the vast majority of attribution seems to be done via 7(b)(1) or (2) (e.g. with hyperlinks).
2. This section is about the contributor of the text, not the subsequent reuser. In other words, if I make an edit -- I (the contributor) agree that these are valid means of attribution. The terms of use generally does not control reusers and in this case this section is specifically not about them.
3. So if text was added by Author A, then Author A agrees that attribution in one of those three ways would be satisfactory. If Author B made comprehensive style changes, Author B also agrees that attribution of their contribution in those three ways would be satisfactory. SSpalding (WMF) (talk) 16:40, 14 April 2023 (UTC)
The section about attribution is very tightly linked with the section about re-use. But I talked about attributing a single section from the article as Jjanhone suggested.
When user B makes style changes across the whole article he makes derivative work. And that work covers the whole article. The question is whether we can list only authors A and B as authors of the section as attribution, or should we list all the authors of the article. D6194c-1cc (talk) 22:41, 14 April 2023 (UTC)

Deepfakes

Would it be possible to prohibit audio (and video) files from being used as samples for voice mimicking? (I have Wikinews Audio and VALL-E in mind.) - Xbspiro (talk) 04:15, 13 April 2023 (UTC)

This is one of a handful of requests related to generative AI that have come up during the discussion. Here's a very recent one that describes our position on this topic in greater detail. To expand on that, we currently think that the terms of use sufficiently handles both violations of project rules and misuses of the Creative Commons licenses.
Although the recent commercialization of has forced many to tackle with the ethics and philosophy underlying open licensing, the current best legal interpretations of the Creative Commons licenses suggest that CC works can be used in ML training as long as the license principles (e.g. attribution) are followed. Said differently, it's not illegal to use audio from the projects to do so.
That said, specific projects like Wikinews may make rules against certain uses as long as they gain consensus and don't violate the project-wide principles of openness and free knowledge. To the extent that they do, then there are already mechanisms to enforce TOU related rules.
Although the widespread commercialization of "deepfake" definitely presents a "new" vector for introducing bias into the projects and will no doubt have an effect on the information ecosystem, it does not present an issue that is novel from a legal standpoint and can be addressed through community rule making. SSpalding (WMF) (talk) 16:23, 14 April 2023 (UTC)
@SSpalding (WMF): Thank you for your answer. To your knowledge, are there any licenses, which would allow a similar usage of material as CC licenses do, but with a ban on using the material as a sample for ML? - Xbspiro (talk) 07:56, 15 April 2023 (UTC)

In the event of any differences in meaning between the original English version and a translation, etc.

These Terms of Use were written in English (U.S.). While we hope that translations of these Terms of Use are accurate, in the event of any differences in meaning between the original English version and a translation, the original English version takes precedence.

In the event of any differences in meaning between the original English version and a translation, the original English version takes precedence.

This line appears twice in the ToU. If you are already updating the policy, maybe just remove the second one and replace the first one with:

These Terms of Use were originally written in American English. While we hope that translations of these Terms of Use are accurate, there may be cases where there is a difference in meaning due to errors in translation. In the event of any differences in meaning between the English version of these Terms of Use and a translation, the original English version takes precedence.

The bold could be kept for emphasis, or some other method of emphasizing the text may be found. If it's really so important that it needs to be kept tacked on to the end, then just make a new section for it like the UCoC enforcement guidelines did. I know the privacy policy and cookie statements could be similarly changed, but those policies aren't being discussed at the moment. –MJL ‐Talk‐☖ 18:08, 14 April 2023 (UTC)

Thanks for pointing this out. It looks like it's mentioned twice since the first time is in the actual body of the terms (the legally operative language that would be used in a courtroom) and the other one at the bottom is meta information (about previous drafts) which is for the benefit of readers but not necessarily useful in court.
I've flagged this suggestion as something to consider, but I think both serve different enough purposes. SSpalding (WMF) (talk) 20:41, 14 April 2023 (UTC)

ทำไมกระผมถึงไม่ได้เข้าสู่ระบบบัญชี

ในเมื่อกระผมทำการลงชื่อเข้าใช้เรียบร้อยแล้วในชื่อ(อนุรักษ์ ศรีจันทรา90)แต่ทำไมถึงบอกว่าผมยังไม่ลงชื่อเข้าใช้บัญชีไม่ทราบครับ 223.24.162.14 23:21, 15 April 2023 (UTC)

กฎบัตรที่ทางท่านต้องการให้กระผมนำเสนอในความคิดของกระผมนะครับ

ขอขอบคุณที่ทาง"wikimedia"ให้โอกาสกระผมในเรื่องของการเขียน"กฎบัตร"ซึ่งเป็นสิ่งสำคัญของการใช้งานโปรแกรมที่คนทั้งโลกยอมรับและช่วยเหลือกันในการนำเสนอหาข้อมูลความจริงต่างๆที่มีกับสิ่งที่เกิดขึ้นบนโลกใบนี้ให้มีที่มาที่ไปหลักฐานอ้างอิงกับสิ่งที่เกิดขึ้นภายใต้ข้อกำหนดของ"องค์กร หน่วยงาน สมาคม กองทุน แหล่งเรียนรู้ที่ก่อตั้งขึ้นภายใต้คำว่า(wikipedia)ซึ่งมีนโยบายหลักของทางองค์กรอยู่แล้วนั้นซึ่งถือเป็นมาตรฐานในการคัดกรองที่เห็นชอบด้วย"คณะกรรมการของทางวิกิพีเดีย"อยู่แล้วกระผมขออนุญาตเสริมตรงที่ทางท่านถามความคิดเห็นมาถึงกระผมดังนี้ครับ "กฎบัตร"(วิกิพีเดีย)โดยนำนโยบายโดยรวมของนโยบายที่มีอยู่แล้วของทาง(วิกิพีเดีย)มาร่วมกับสิ่งที่กระผมได้เขียนขึ้น3ประการและหนึ่งหัวใจดังนี้ 1.นำเสนอ คือ การสร้างขึ้นที่มีการกล่าวถึงที่เป็นจริง 2.กล่าวอ้าง คือ หลักฐาน ข้อมูลการกระทำถึงสาเหตุที่สามารถอ้างอิงได้และน่าเชื่อถือโดยเป็นที่ยอมรับโดยสากลโลก 3.แบ่งปัน คือ การกระจายความรู้สู่ทุกระบบและผู้ต้องการค้นคว้าและอ้างอิงข้อมูลที่ถูกทำการตรวจสอบแล้วว่าถูกต้องตามปัจจุบันก่อนที่จะมีเหตุผลที่น่าเชื่อถือและอ้างอิงได้ในอนาคต มาถึง"หัวใจสำคัญของกฎบัตร คือ (จรรยาบรรณ)"ของผู้นำเสนอ ผู้ตรวจสอบ ผู้ปฎิบัติ ผู้ถ่ายทอด ผู้สืบทอด และผู้นำไปใช้ควรมีซึ่ง"จรรยาบรรณ"ของตนและเคารพรักษาสิ่งสำคัญนี้ไว้เสมอครับผม 

ผู้เรียบเรียง นาย อนุรักษ์ ศรีจันทรา

16/04/2023เวลา6:54น.ตามเวลาในประเทศไทย 223.24.162.14 23:55, 15 April 2023 (UTC)

Third Party Resources Policy

Section 4 — Please add a suitable link to the currently applicable Third Party Resources Policy — GhostInTheMachine talk to me 09:42, 16 April 2023 (UTC)

Already referenced in #Third Party Resources Policy above — GhostInTheMachine talk to me 10:16, 16 April 2023 (UTC)
Based on feedback, we have made the decision to cut its inclusion in the next draft of the TOU since the specific policy language may not be completed during this consultation. SSpalding (WMF) (talk) 16:47, 16 April 2023 (UTC)
Good. Thanks — GhostInTheMachine talk to me 16:52, 16 April 2023 (UTC)

Another minor copyedit/nitpick

In the "termination" section, including Sections 1, 3, 4, 6, 7, 9-16, and 18 should read including Sections 1, 3, 4, 6, 7, 9–16, and 18 (using an en dash instead of a hyphen). Best, HouseBlaster (talk) 17:03, 16 April 2023 (UTC)

General dislike for the mediation section

Earlier discussion

I dislike this section as it applies to a new editor anonymous editor who knows nothing about paid editing, and is editing on controversial topics.

Some thoughts:

  • Clauses fundamentally affecting people's assumptions about legal rights should probably be top level clauses rather than hidden in another section.
  • I don't think you should talk about "mediation" at all, but should use the term "binding arbitration", since the term "mediation" seems to be used for non-binding processes so the term is apt to mislead, and you should be clear about interfering with people's access to the courts - so should emphatically refer to "binding arbitration" (See this company blog post)
  • As a point of fairness, I would be concerned that wikimedia would pick some biased kangaroo court. The terms do not specify how the arbitrator would be chosen - in fact from reading this it sounds like the WMF could be arbitrating cases themselves.
  • If wikimedi started demanding documents from me as an anonymous paid editor my reaction would not be very positive. I imagine the same is true of a number of anonymous editors editing on controversial topics.
  • Using your own parallel court system to work arounds the norms about court costs in your country is a little dodgy.

I guess this is how "lawfare" is played... but is it actually worth playing games with access to the courts and presumptions of anonymity in order to address paid editing.

Talpedia (talk) 19:59, 16 April 2023 (UTC)

Seriously? We can expect paid editors to make themselves familiar with our rules before they edit any Wikimedia projects, that's what they get paid for. And if paid editors don't like the mediation thing, they could simply a) make sure they comply with our rules or b) stop paid editing at all.
Paid editing is a huge burden for volunteers, cleaning up the mess which marketing companies often deliberately create. It's great that the WMF finally steps in to take action in case of paid editors violating the ToU. Johannnes89 (talk) 20:12, 16 April 2023 (UTC)
Yes @Johannnes89: I am serious, that's why I said. You seem unaware of the facts that rules apply to everyone including 10 edit newbies. Talpedia (talk) 20:45, 16 April 2023 (UTC)
The paid editing rules do not apply to newbies, but only to paid editors. Paid editors are expected to carefully make themselves familiar with our rules and properly disclose their paid editing. If they fail to do so, they are no innocent „10 edit newbies“, they could and should have known better, that's what they get paid for. Johannnes89 (talk) 06:05, 17 April 2023 (UTC)
Perhaps I don't understand, but I thought that every wikipedia user would be required to agree to these terms, allowing the WMF to sue them in their own court with its own process if it suspects them of paid editing (as the wording stands) Talpedia (talk) 14:01, 17 April 2023 (UTC)
The ToU apply to all users, but the marketing company mediation of course only applies to paid editors (and in practice only to those getting caught). WMF is not looking to sue individuals but rather companies violating our rules time and time again, such as this one [11][12][13] Johannnes89 (talk) 15:55, 17 April 2023 (UTC)
Yes, the most relevant thing I would add for Talpedia is that this is targeted toward marketing companies that intentionally violate local project rules through systematic undisclosed paid editing campaigns.
So "innocent" editors who don't know the rules, people writing on controversial topics, etc are not within the scope of what these changes are meant to address. This is meant to stop bad actors who are paid by companies to engage with the projects to waste volunteer contributors time. It is further meant to get the companies to stop, not necessarily punish individual editors.
Finally, the action is a mediation (a negotiated settlement). The action only becomes an arbitration should the party be unwilling to settle. I went into additional detail in other discussions here, but the terms of the negotiated settlement would be relatively clear, "stop violating the project rules." It would only turn into an arbitration should the bad actors refuse that request. It's included because this gives us a way to enforce a settlement like this in a court should the company continue to violate the rules. SSpalding (WMF) (talk) 16:32, 17 April 2023 (UTC)
I mean that's all well and good... but if it's not specified in the contract it all amounts to "trust that we won't use the contract in the way we say we can" e.g. that WMF will only identify the correct individuals. Why should people trust the WMF, particularly when it decides to exempt itself from the courts? I guess at least there is the restriction to paid editing and it is fairly easy to avoid such editing - though perhaps less easy to not be misidentified as a meat puppet engaged in it. Not sure I follow the whole "mediation unless you refuse to settle" - it's sort of "agree or we'll force you to agree" which very much seems like binding arbitration to me, though perhaps there are some details that make it different in some material way. This line of arguments comes up in forced medical treatment a bunch "take the drugs voluntarily or we'll imprison you hold you down and make you", I really dislike the argument there, to the extend that I dislike similar forms everywhere. Coercion should be clearly stated. Talpedia (talk) 22:36, 17 April 2023 (UTC)
There are some big picture things that might help allay your fears of arbitrariness. Most importantly, there are, objective bright-line rules: if a company "discloses" the payment it gets for contributions according the the rules of the local project, then that is all they need to do to avoid violating this term.
I'd argue that this is not like "forcing a person to take a medication to avoid imprisonment" for many reasons. Chiefly in this situation, there is a totally viable alternative. If a company does not want to comply with disclosing their monetary interests in the way the community asks them to (because they want to intentionally introduce bias into the projects and create more work for volunteers), then that company should simply not be allowed to get the benefits of the community. The company's bodily integrity isn't being compromised. Its freedom is not being compromised.
Sure, it's different in many ways. My point is more that I recognise the harm of hidden coercion in a few places, so wish for it be avoided elsewhere - I don't think that being clear about coercion (such as by using the title Binding Arbitration rather than mediation) has particular downsides. In this statement, I'm complaining about the hidden part. If we want to force people to accept quasi-legal settlements that they can't appeal in our own courts then maybe that's okay but we shouldn't pretend that people have "declined to settle" or that there has been mediation: we've summoned them to a hearing where we tell them what to do and then if they don't do it we impose a punishment. Now in practice the pseudo-courts we use may be fair and emuluate the real courts well or even provide superior justice, and this might all be contained within case law surrounding the term arbitration and the act specified, but that's not immediately clear from the wording. Talpedia (talk) 07:13, 18 April 2023 (UTC)
I realize that you may see arbitration itself as something that you fundamentally disagree with, so I am not trying to argue with your philosophical stance. But as a practical matter, we are trying to make it more clear to companies trying to take advantage of volunteers that we have a viable method of taking them to task. We already do have the power to litigate against these companies since 2016. But it's difficult for me to argue that forcing a company into a six-figure ($) year-long litigation is more fair and humane than a four-figure, one-day mediation. SSpalding (WMF) (talk) 00:13, 18 April 2023 (UTC)

But it's difficult for me to argue that forcing a company into a six-figure ($) year-long litigation is more fair and humane than a four-figure, one-day mediation

The issue is with the fairness and oversight of the process. I fully agree that the legal processes impose a lot of costs on those involved. But the same argument that you quote here applies to, say, same day martial law. What real courts add is oversight and democracy.
Even in the presence of arbitration there are a few practical issues: is it clear from a cursorary skim of the contract that an individuals legal rights are fundamentally being altered? Which arbitrator are we using? How is the arbitrator regulated? What are the arbitrators rules for evidence and process? Is the person signing the contract able to assess the arbitrators reputation for justice? Does the arbitrator follow the US law or does it have it's own law (as is the case for sharia law arbitration in the UK)? What is the cost of the arbitrator? Are its awards capped?
Would you sign a contract stipulating unbounded liability for your counterparty's losses assessed solely by your counterparty? If not, how is this different - the main difference is the presence of the arbitrator. The impositions that the state legal system apply are onerous but at least they are shared, regulated and controlled democratically, none of this necessarily applies to a quasi-legal process (outside of judicial review). Talpedia (talk) 07:13, 18 April 2023 (UTC)
As you are talking about fairness: What's unfair is the advantage in time and resources paid editing companies have over volunteer communities. It takes a lot of time to clean up their mess and all they have to do is create new sockpuppets and continue.
If they choose to violate the ToU time and time again, it's certainly not unfair to establish an easier way to legally stop them from continuing their abusive behaviour.
„Would you sign a contract...“? They don't have to sign the contract, they could just stop paid editing if they don't like our rules. And if they want to continue paid editing at Wikipedia they can simply choose to follow our disclosure rules and they are fine. Johannnes89 (talk) 14:34, 18 April 2023 (UTC)
Again... I'm still concerned about the average user rather than the paid editor. I think the difference is that you trust the WMF to never make mistakes, perhaps I am optimising for a possibility that will never happen. To use an analogy that is perhaps a bit excessive, it's as though someone include a clause "all shop-lifters arrest by the police will be found guilty without trial and have their hands cut off" and the response to concerns about process were "shop-lifting is a terrible crime and something must be done about and the courts are too slow." We want to either be satisfied that the arbitration process is and will always be "fair enough" or that WMFs identification of paid editors will always be "correct enough" Talpedia (talk) 16:40, 18 April 2023 (UTC)
There's also another problem with paid editing. In my country, if a worker write texts by official tasks from employee, copyright is transferred to the employee (also contract must cover payment for the transfer). D6194c-1cc (talk) 10:04, 17 April 2023 (UTC)

18 April 2023: Second update notice

Hi all. We have updated the Proposed Update again (comparison of the versions) and posted an explanation about it, based on the feedback from this talk page and from the two office hours. You can still add your thoughts and feedback on this version of the Proposed Update below this message. Thank you very much for your participation.

On behalf of the Legal Department,
RamzyM (WMF) (talk) 08:05, 18 April 2023 (UTC)

Section Paid contributions without disclosure

ToU: "A Wikimedia Project community may adopt an alternative paid contribution disclosure policy which may supplement or replace this section. If a Project adopts an alternative disclosure policy, you may comply with that policy instead of the requirements in this section (titled “paid contributions without disclosure”) when contributing to that particular Project."

Question: in spanish Wikipedia, inside the COI policy there is a section about paid contributions. In this case, would it be considered an alternative paid contribution disclosure policy? There is difference between creating a policy only for paid contributions and having a section inside COI policy. Thanks Cartago90 (talk) 11:12, 19 April 2023 (UTC)

You asked a similar question here, "Alternative paid contribution disclosure policy" that I provided our best answer to. To summarize, if there is an enforceable local rule about paid contributions, then those rules with supplement and amend the interpretation of the TOU. The TOU is meant to be useful for Spanish Wikipedia to help that community enforce the rules about disclosure it wants people to follow. It is not there to override rules that it has chosen to have locally and create problems where a community feels like there aren't any. If you feel like there is a grey area where the phrasing is unclear, then here's a rule of thumb. If something breaks a rule that Spanish Wikipedia believes should be enforced, assume that it also violates the terms. If it does not break any rules of the local wiki or project-wide rules, then assume that it doesn't violate this portion of the terms. SSpalding (WMF) (talk) 18:37, 19 April 2023 (UTC)

Two criticisms

Paid contributions without disclosure

You must disclose each and any employer, client, intended beneficiary and affiliation with respect to any contribution for which you receive, or expect to receive, compensation. You must make that disclosure in at least one of the following ways: a statement on your user page, a statement on the talk page accompanying any paid contributions, or a statement in the edit summary accompanying any paid contributions. In addition, if you make a public posting off on the Wikimedia Projects a third-party serviceadvertising editing services on Wikipedia in exchange for compensation of any kind pay, you must disclose all what Wikipedia accounts you have used or will use for this service in the public posting on the third-party service. Applicable law, or Project specific policies and Foundation’s policies and guidelines, such as those addressing conflicts of interest, may further limit paid contributions or require more detailed disclosure. For example, infringing Wikimedia Trademarks (defined hereinafter at Section 6) to advertise paid editing services, removing disclosures on previously adequately disclosed paid edits, or logged-out paid editing in a manner that makes adequate disclosure a practical impossibility violate this section.

A Wikimedia Project community may adopt an alternative paid contribution disclosure policy which may supplement or replace this section. If a Project adopts an alternative disclosure policy, you may comply with that policy instead of the requirements in this section (titled “paid contributions without disclosure”) when contributing to that particular Project. An alternative paid contribution policy will only supersede these requirements if it is approved by the relevant Project community and listed in the alternative disclosure policy page.

It is of my personal opinion that all paid edits should be banned, period. It’s very questionable whether paid edits can ever satisfy NPOV and policies on conflict of interest. There also are immediately questions of financial interest involved when money’s at the table. These are problems regardless of intent and requiring disclosure does not make those problems magically go away. Problems arising from paid editing have been noticed by e.g. journals before.

We reserve the right to suspend (temporarily, or permanently) our handling of reports or other correspondence from users or third parties, whether about allegedly illegal or otherwise problematic content or conduct, or requesting appeals against moderation actions, if such correspondence was made in bad faith, repetitive, unfounded, and/or abusive. In appropriate circumstances, your email address may even be blocked on our email system(s), and you will then need to contact us at our postal address if you wish to further correspond with us during that block. For less serious cases (e.g. up to three polite emails about one or more meritless complaints), this is likely to be temporary. More frequent or more abusive communications are more likely to lead to permanent measures.

The definition of abuse here is very subjectivist and allows Wikimedia to bar people from reporting for basically any reason. Also what’s the point of blocking email addresses? Iota Mendeso (talk) 21:44, 20 April 2023 (UTC)

Hello, Iota Mendeso. In terms of your first point, I wonder if that is voiced from a specific Wikipedian perspective. For instance, if a library decided to pay staff to archive content to Wikisource, would you regard that as also problematic? If a school used Wikiversity to create course materials, that might not constitute much of a conflict of interest - rather it might be mutually advantageous. The Terms of Use do explicitly permit individual projects to supersede with firmer restrictions if required. In terms of your question about the point of blocking email addresses, speaking from my perspective as the VP overseeing Trust & Safety and Human Rights, we unfortunately see some people who abuse email access rather egregiously. The open postal address permits people to reach out to the Foundation without being able to, for instance, communicate threats or hate speech directly to random staff members. It can also help prevent email bombing. --Maggie Dennis (WMF) (talk) 14:52, 24 April 2023 (UTC)
It is a problem if the staff do not disclose exchange for compensation, money or gifts appeal to UPE more than improve encyclopedia, they can make trouble stubbornly to disrupt Wikipedia, no matter advertising or other kind, so I think they must disclose all paid editing. Long5563 (talk) 08:23, 30 April 2023 (UTC)

Repeating my request on other countries' laws

Could we please clarify that violating the laws of a country other than the US is not a ToU violation, and state that the Foundation will resist efforts to prosecute such charges? 2601:647:5701:39B0:E4A5:29B1:E1FB:6A77 — The preceding unsigned comment was added by 2601:647:5701:39B0:E4A5:29B1:E1FB:6A77 (talk) 14:56, 23 April 2023 (UTC)

see replies at Talk:Terms of use#Responsibility for laws where subjects written about are located?, this is meant as a warning, that other countries might trie to take legal action against editors. --Johannnes89 (talk) 15:54, 23 April 2023 (UTC)
I understand that, but I still feel that the clarification is necessary. Those editing on topics disfavored by despotic regimes need to know that they will not be in trouble, and that the Foundation will try to shield them. 2601:647:5701:39B0:E4A5:29B1:E1FB:6A77 16:58, 23 April 2023 (UTC)

If this can't be done, can we at least specify that the Foundation isn't interested in violations of country's laws other than the USA? How much can we say about what the Foundation will do to shield editors from non-USA charges? 2601:647:5701:39B0:A652:FC3B:B355:8C65 21:39, 30 April 2023 (UTC)

Tasks of the community - UPE - Marketing Company Mediation

As it has been said in the above section Marketing Company Mediation, the new process won't change the need for good-faith editors to identify UPE, revert their edits, and potentially debate with these bad actors.

So which burden of the editors exactly will this process of Marketing Company Mediation help relieve ?

Will you publish somewhere the new do's and don't of the Community when the Marketing Company Mediation will be enforced ?

Is there any risk of interference of the community in the process of Marketing Company Mediation that UPE could use to skew the mediation ?

Will there be warnings on Wikipedia about scams ? Will you give to the readers of Wikipedia the possibility to report or complain about scams or promotional content ?

Thanks Waltercolor (talk) 16:44, 23 April 2023 (UTC)

Thanks for your feedback. It is accurate to say that nothing about the update changes the way that editors will identify UPE and deal with it day-to-day.
The update is meant to provide new opportunities for the Foundation to help when those processes aren't enough: for example, when a company engaging in undisclosed paid editing makes dozens of hundreds of sockpuppets. They have the resources to do this because they are financially incentivized to introduce problematic edits. Admins can only deal with them *after the socks violate the rules of the project*. As a result, admins have to play "whack a mole" with well-financed, dedicated "moles."
So for a concrete example, should a community investigation identify a set of coordinated bad actors, the Foundation's Legal department can intervene with a court-ordered mediation to get them to stop what they're doing currently and in the future. And if they choose not to follow the rules, we can legally be entitled to penalize them financially for the damage their cause. This would limit their financial incentive to continue their rule-breaking and hopefully, encourages them to move on to a less risky way of making money off of the projects. For various reasons discussed in other sections, threatening litigation (which is currently the only thing that we can do) isn't as credible of a threat as actually engaging in the mediation (which this allows us to do).
Interference - this is discussed elsewhere in greater detail. In sum, the community can clean up undisclosed paid editing without fear of interfering with any potential case we have against the marketing company. All of the information needed to engage with mediation is on wiki: for example, the history of edits on a page.
This is also a jumping-off point for some practical approaches as you suggest: the new rule gives us an opportunity to communicate expectations to marketing companies that have already been causing problems. It's also an opportunity to post best practices for companies editing the projects and for vendors who are offering services related to the projects. We have posts like this in the works, but I can't confirm exactly what will be published and when until they've been through review.
Readers and victims of companies do periodically reach out to the Foundation's Legal department about scams and UPE. A lot of what we learn is through community investigations that are already on Wiki, so that is also a good place for editors to let us know what's going on. Although is unlikely to be an email address specifically for UPE reporting, legal-tm-vio@wikipedia.org is an address that is reviewed regularly by the legal team where UPE violations have been reported in the past SSpalding (WMF) (talk) 17:19, 24 April 2023 (UTC)
Thanks a lot @SSpalding (WMF) for these useful answers. And sorry for "crossposting" the question at the Live Conversation of the 27 April about the Annual Plan. @Jrogers (WMF) was providing nearly the same answers and I thank him for the clarification.
I realize it's good to have such conversations cause it give us all more insight about our different perspectives.
Your policy and answers to the problem of UPE is focused on the harming of projects.
But my question was about harming persons. The agencies hiring UPE are actually targeting our volunteer goodfaith editors.
Wikipedia is constructed, moderated and modified by a community of volunteers.
You wisely described UPE not as "editors" but as "users". But for the so called "good faith editors", it's crucial to know with whom they collaborate and to detect UPE, because it's not really possible to build a safe community with both neutral volunteer editors and marketing-driven paid editors (in fact, these are not just "users", they edit content and are not neutral).
So I believe the UPE question should not only be treated as a legal WMF problem as you suggest, but also as a threat for the community itself. We should evaluate regularly the proportion of undisclosed paid editings in the editing process and undisclosed paid editors amongst the persons editing and see how these two figures evolve, if they increase, remain stable, or decrease.
So I suggest that additionnaly to the legal aspects which will be implemented like MedArb, the WMF and the communities meet together regularly in, say, a UPE yearly Conference about the topic of Undisclosed Paid Editors and editings.
This would be a moment when the community and the WMF could share infos and thoughts about their specific concerns with UPE and evaluate regularly the dynamics of this phenomemon.
Also keep in mind that the WMF and marketing agencies function both with paid persons, which can be replaced if necessary.
It's not the case for volunteers in the projects. If the good faith volunteers quit the movement because they are exhausted, they will be much harder to replace than paid editors.
Under the pressure of marketing agencies, some volunteer editors may also be tempted to become paid editors.
If a growing proportion of experienced or new editors on Wikipedia are paid editors and no more "goodfaith volunteers", this could change significantly the ecosystem of free sharing of knowledge.
We should at least monitor it. If we do not cross our infos, we may not be aware of this.
Waltercolor (talk) 09:00, 29 April 2023 (UTC)

27 April 2023: Terms of Use update process closed

Hi all,

The update process has ended. Please find the following message from Associate General Counsel Jacob Rogers (originally posted at wikimedia-l):

Hi everyone,

The comment period for changes to the Terms of Use has closed as of April 24, 2023. On behalf of the Legal department and the Foundation, I’d like to thank everyone who offered questions, comments, or even just stopped by to read through the page out of curiosity! We were pleased with the quality and volume of comments. They have been integral to creating a Terms of Use document that will serve the community and the projects going forward. Even when some feedback could not be directly integrated in changes in the terms of use text, we’ve still reviewed your comments and they have helped influence the Legal department’s other policies and work.

We have shared a couple of draft iterations during the 60 day comment period. The next step will be for the legal team to take some time to review everything, draft a final version that will be published in mid-May on the Meta consultation page[1], then we will finalize the approvals and put the update officially on Foundation wiki.

Thank you to everyone who provided questions and comments!

Best,

Jacob

[1] https://meta.wikimedia.org/wiki/Wikimedia_Foundation_Legal_department/2023_ToU_updates

Best, RamzyM (WMF) (talk) 10:48, 2 May 2023 (UTC)

@RamzyM (WMF), @SSpalding (WMF): and @Jrogers (WMF) - just thought I'd offer a big thumbs up for how the ToU consultation process has been handled. Positives include:
  1. Increased explanation of why each change was made;
  2. Rapid responses to most feedback/questions;
  3. Flexibility in the zoom calls in the topic;
  4. Open stating of what were the recurring requested changes, even when they weren't being implemented as amendments
  5. Explanations of why those amendments weren't used
  6. Actually making changes
Mark it in the box for methodologies for future ToU consultations, with thanks for all in the team Nosebagbear (talk) 11:09, 2 May 2023 (UTC)
+1, agree with every word Nosebagbear said. Johannnes89 (talk) 18:34, 2 May 2023 (UTC)

May 22 Final Draft Comment

Hi everyone,

As we've posted our final draft of the Terms of Use, we wanted to share a note of thanks to everyone that participated in the conversation!

The talk page had over a hundred conversation topics and we held two office hour conversations covering a variety of issues that go to the core of how people are able to safely access and contribute to the Wikimedia websites. As a few examples of how this influenced the draft, community comments helped us improve the language for applicable laws to clearly say that the Foundation won’t help enforce laws that violate the human rights of users. Comments also helped us improve the ToU as a document including better parsing for readers with an audiovisual impairment who use a screen reader and improved language in several spots to help with translation in the future.

As next steps in the process, we are sharing the final draft of the ToU here to Meta and with the CEO and General Counsel of the Foundation for review and approval under the 2016 delegation of policy-making authority. We have also informed the Board about the update. When the formal approval is finalized, we will post the updated terms to the foundation.wikimedia.org site. We will also be working on updating the draft translations for the consultation languages to the final text, with additional language translations to follow in the next several months.

Thank you again everyone for the thoughtful questions and comments throughout the discussion! -Jrogers (WMF) (talk) 17:16, 22 May 2023 (UTC)

You can find the comparison of the previous version with the final version of the proposed update here. RamzyM (WMF) (talk) 06:11, 23 May 2023 (UTC)
Retrieved from "https://foundation.wikimedia.org/w/index.php?title=Policy_talk:Terms_of_Use/Archive_2&oldid=344480"