Draft feedback

I started a draft. I feel pretty good about it, although there are still a couple placeholders. I don't think we're going to find a consensus for any enforcement, but perhaps we can come to a consensus about a guideline. I've tried to define copyleft trolling, give some tips to distinguish it from other types of enforcement, explain what I think there is strong consensus for in terms of principles, and detail some of the ways we've dealt with it in the past. The point isn't to provide anything prescriptive, but to document "this is something we care about and could do something about".

Pinging some of the users active in the VP discussion: @Nosferattus, SHB2000, King of Hearts, Normanlamont, Adamant1, Jeff G., Chris.sherlock2, Jmabel, Nil Einne, Yann, Enyavar, Bluerasberry, and Julesvernex2: (sorry if I missed anyone -- I just pinged those whose name stood out more than once in that thread) — Rhododendrites ^talk |  15:35, 28 May 2024 (UTC)

• Thanks Rhododendrites, this is a great start. I think the 'Victim of copyleft trolling?' section is a key one, as many users will likely land on this page only after being hit with a claim. I've reached out to CC for their feedback on which jurisdictions/scenarios Pixsy & Co. are more likely to pursue legal action, and what actions victims can consider to decrease the odds. As you said, we need to fall short of providing legal advice, but hopefully there's some useful stuff we can include. --Julesvernex2 (talk) 17:25, 28 May 2024 (UTC)

  Great start! I agree that we'll need to fall short of providing legal advice. Maybe we should also point out the fact that CC-4 gives a 30-day grace period? SHB2000 (talk) 21:32, 28 May 2024 (UTC)

  @SHB2000: That's in there, but remember that grace period only applies to restoring the license. It doesn't stop the demands for money; it only means you can continue to use the image once you've paid. With the older licenses, if you pay a fee, fix attribution, and continue using the images, you could be sued again because you didn't have a 30-day window to restore the license. It's not much of a fix, but it does reduce some of the harm. — Rhododendrites ^talk |  22:17, 28 May 2024 (UTC)

• This is a nice start. Thanks for your work on it. So far I've been rather disheartened that it seems almost impossible to build consensus to do anything about copyleft trolling here. Maybe this page will help to inform people so that we can start to build some consensus around defending our values and protecting reusers. Nosferattus (talk) 03:51, 29 May 2024 (UTC)

• Thanks all. I'm going to move this into projectspace and remove the essay template (without replacing it with anything). Here's my question: should this be shaped into something that can be proposed as a formal guideline? Or should it be a "help" page? It's explaining and articulating some principles rather than prescribing any particular behavior, which makes me think it could be out of place as a guideline, but maybe being a guideline would give those principles more weight? Curious about your thoughts. — Rhododendrites ^talk |  13:27, 29 May 2024 (UTC)

  Thank you for doing this. I’m sure it will help a lot of people. Normanlamont (talk) 16:56, 26 June 2024 (UTC)

@Rhododendrites: I like it, but I had to add a word. I suggest making it a guideline.   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 12:56, 31 May 2024 (UTC)

Finally took the time to fully read this. It is a great overview on the current situation, and a solid foundation to build on once the community decides on further actions. --Enyavar (talk) 15:38, 12 June 2024 (UTC)

Negotiation

@Nosferattus: Regarding this, to me the line you removed is less telling people to pay trolls and more telling people they don't have to pay the amount stated, which seems useful to know. What about modifying that line to start with "If you decide to pay the fee..."? Certainly we shouldn't tell people not to pay, as ignoring the demands can turn accidental infringement claims into "willful infringement". — Rhododendrites ^talk |  14:19, 29 May 2024 (UTC)

That sounds reasonable. FWIW, I do think ignoring the monetary demand is an option and we should not preclude that. Nosferattus (talk) 17:22, 29 May 2024 (UTC)

Definition

I think the article needs to have something in the lead about valid licence enforcement (rather than half way down) including your comment about otherwise the image would be functionally public domain. The lead sentence mentions "for minor licensing violations". The current Diliff case was brought about by a user who admitted they didn't read the licence conditions at all, thought the image was public domain, and didn't attribute Diliff at all. That's not a minor licence violation. They essentially treated the image as "If it's on the internet it must be free". And we also don't know anything about this person. They could work in the advertising dept of a large corporation for all we know. So in fact it might not fit into the model described and be indeed a perfectly reasonable example of licence enforcement.

The main issue for which we actually have facts, is that Diliff has employed Pixsy for licence enforcement. What is definitely different about the Diliff case is there was no prior intent to upload images from which an income stream would arrive as Dillif was active here for years. Diliff has himself expressed that he has tried to intervene to help out the less egregious misuse and personally has no desire to demand big fees from individuals or tiny businesses.

I think this is more of a case of where Pixsy has been used as a blunt tool for licence enforcement of misuse by large corporations, with collateral damage. Can we distinguish that somehow? The idea that Diliff personally wishes to seek large fees from minor licensing violations from individuals and small businesses is false, as Diliff has explicitly said that is not their intent. Really, we need something saying please don't use Pixsy for licence enforcement. -- Colin (talk) 09:00, 30 May 2024 (UTC)

@Colin: Let us consider either rearranging file description pages to put the licensing first, or putting a "Licensing" link (in the appropriate language, to the licensing selection further down) above the file display. I don't know how feasible either would be.   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 13:01, 31 May 2024 (UTC)

So proposed at COM:VPP#Make licensing easier for reusers to see.   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 13:20, 31 May 2024 (UTC)

I've been reading more on this. The Pixsy ban proposal, which failed had useful input from one user who made use of Pixsy while claiming to be behaving honourably. They made use of a template User:Josve05a/Pixsy that appeared in their image description page that warned users the image was protected against "image theft" by Pixsy. This might be a useful pattern to follow, though the Media Viewer doesn't pick it up.

The CC Enforcement principles do not mention anything about "differentiate between independent/small-time reusers and large corporations" which this page does. I'm getting the feeling that this is more of personal viewpoint than something that really should appear on this page. I don't think any enforcement practice could request information on company turnover and even independent and apparently small firms can have turnovers or routine business expenses for which a few hundred quid is a minor trouble. The article later notes that Pixsy only pursues commercial re-users. I suggest this be reworded to say trolls go after private individuals and non-commercial operations.

CC's licence enforcement page says "We believe it is important that reusers make their best efforts to comply with the terms and conditions of the license, and we support enforcement of the licenses when CC-licensed materials are being used without regard to those simple terms." Note this "without regard to those simple terms". I think an important distinction is where there has simply been "image theft", that the image has been taken and reused "without regard to those simple terms". This is quite different to when the attribution is simply wrong in some way. I would be interested to know if Pixsy's search is cleverer than just an image search but seeks out attempts at attribution such as mention of the photographers username or real name. CC say "The CC licenses reflect a deep-seated belief that authors deserve credit for their work, consistent with moral rights in force around the world." Clearly CC see a difference between a good faith attempt that didn't quite reach the mark, and complete disregard and disrespect for the content creator.

The CC Enforcement principles demonstrate best practice. But we are a broad church and there's a gulf between saintly behaviour and unacceptable behaviour. They highlight "not a scheme to trap" and "When enforcement becomes profitable, especially when licensors act in a way that suggests they want reusers to violate the license so they can collect fees, it crosses over into trolling and runs counter to these principles." I think there is a level where enforcement is not compliant with CC's own (quite recent) enforcement principles, but has not reached the level of copyleft trolling. I think that line gets crossed when there appears to be a prior intent to create and upload works with an apparently free licence that then traps reusers making minor attribution mistakes, targets any and all non-compliant reusers including non-commercial reusers, and which operates at a scale that is a "business model". This line could be the kind of hard line around which we developed a policy for deletion of such images and banning such users. Below that line is behaviour that we don't regard as best practice and doesn't comply with CC's enforcement principles. I suspect there could be a wide range of opinion among professional photographers of the application of forgiveness and spending one's time helping commercial users comply that would appear to load that "cost" onto the photographer rather than onto the reuser business. One spends an evening helping a chain of BMW car dealers comply with the CC terms, on an image they reused "without any regard to" the photographer, and get diddly squat in return. In contrast, in our broad church, are creators who either explicitly give away their works without strings, or attach strings merely in the hope that some will comply but not care when some don't. This range of attitudes is important to Commons as we take images not just from Commoners and not just from saintly Commoners, but scraped off the internet from people we don't know, don't share all our values and whom we can't actually ban.

The core of CC's definition of copyleft trolling is "licence enforcement as a business model". As with any business model, it requires an intention to lure "customers", to require all "customers" to pay, and to operate at scale. Being non-compliant with every single one of CC's enforcement principles is not quite enough to reach what I believe they mean by "copyleft trolling" but it isn't good either. I think there could be an argument that FP and QI require either a CC v4 licence or PD or CC0. I suspect that would probably rule out a huge number of images taken from other websites such as Flickr, which AFAICS, still use CC v2. -- 12:08, 15 June 2024 (UTC)

Wikicon proposal

Just a heads up to those interested, after digging into this quite a bit over recent weeks I decided to propose a session about copyleft trolling at this year's WikiConference North America. I framed it around poka-yoke, or "error-proofing". (i.e. talking about what the issue is, but focusing on how to prevent it rather than how to squash it). I plan to gather some of the ideas floated here and at the VP, put some visuals together for them, and hopefully gather some ideas from the big crowd who attends there. — Rhododendrites ^talk |  03:36, 1 June 2024 (UTC)

I could help raise awareness at the German Wikicon this October, probably in one of the (shorter and less formal) lightning talks. Right now there is also time for topic proposals for the official agenda. --Enyavar (talk) 15:38, 12 June 2024 (UTC)

Belated follow up on this. The presentation went forward, with a good sized audience and positive feedback, but it doesn't seem to have motivated anyone to come participate on this page. Ah well. For anyone interested, it's here. (Probably shouldn't be uploaded to Commons as there are a couple fair use screenshots in there). — Rhododendrites ^talk |  14:55, 22 March 2025 (UTC)

VPP and DR closures

The two long threads that sparked discussion on this subject have now been closed.

On June 24, IronGargoyle closed Commons:Deletion requests/Files uploaded by Diliff as

Kept: A clear majority here support keeping the images in question given their high quality and within-scope status. Commons policy does not provide a justification for mandating the deletion of these images. There is no consensus about forced watermarking, but it seems discussion is ongoing at the Village Pump to determine how best to deal with similar cases in the future.

Today Matrix closed Commons:Village pump/Proposals#Prohibit copyleft trolling with this statement:

This discussion has run its course, and I am hence closing it. Out of these proposals, it is clear numerically that proposal no. 4 is the most successful.

• The help page Commons:Copyleft trolling has been created, and subsequent comments can be made on its talk page.
• Idea is we can convince users to switch to CC4, and then they can continue to use firms such as Pixsy.
• Authors who still do not switch to CC4 should accept forced watermarks so proper warning is given to reusers.

Per the successful proposal, I am also adding the template to make Commons:Copyleft trolling an official guideline on Commons.

Both of these discussions had some problems whereby earlier participants were operating according to incomplete information. In the DR, the comments after Diliff finished responding are a bit more complicated than the row of keeps at the start. Given the complexity but preponderance of keeps, IronGargoyle sensibly closed as keep while pointing to the Village Pump thread for more general guidance.

In the VP discussion, the three bulletpoints Matrix listed are based on the three bulletpoints in the proposal, which were worded before we have complete information about CC4 and Pixsy. Namely, (a) CC4 doesn't require that reusers be given a chance to correct a violation to avoid demands for money; it provides a window for reusers to restore their license to avoid being sued again after paying those damages. i.e. It doesn't protect against someone seeking damages the first time. (b) Per my conversation with Pixsy, they simply do not deal with images licensed with CC4. They may just think it's too legally blurry or they may want the ability to go after people more than once, but my sense is they're operating under the same mistaken assumption about CC4 that we were, so it's possible they'll modify the policy in the future.

So I'm mainly just hoping to get IG and Matrix to share their thoughts about the discussions the other closed and ask what their recommendation is, following those, to modify this page. — Rhododendrites ^talk |  13:26, 30 June 2024 (UTC)

I closed the VP discussion because it was stale (last comment was 21 days ago) and the most successful proposal was number 4. I think Julesvernex2's comment provides some insight into why CC4 is still better. The most important part IMO was, "most courts will hesitate to award material damages for temporary infringements. As long as there are CC2/CC3 licences out there, Pixsy has bigger fish to fry". The conclusion we can draw from this is that CC4 is still the better choice and we should still encourage its use over CC2/3. Is the limitations of CC4 a thing that should be at least mentioned in this help page? Absolutely. Does it change the crux of this proposal? No. That's my personal take at least. —Matrix(!) {user - talk? - _uselesscontributions} 17:28, 30 June 2024 (UTC)

@Matrix: Thanks. One more question for you. You closed with Authors who still do not switch to CC4 should accept forced watermarks so proper warning is given to reusers. - given that you've also added the guideline template to this page (and guidelines have some degree of enforceability), how would you word that bit for inclusion here? Or perhaps no change? Right now this page primarily explains the concept of copyleft trolling, tries to list some ways it can be differentiated from other forms of enforcement, explains why it's contrary to the spirit of Wikimedia/Commons, and offers some ways Commons has dealt with it in the past. i.e. it's not prescriptive. Is that consistent with your closure? — Rhododendrites ^talk |  17:36, 30 June 2024 (UTC)

The current page does indicate solutions on the section "How Commons handles copyleft trolls". Notice the use of the word "should" in my summary. Ultimately this is a very messy problem, and there is no "one size fits all" solution. This is something that should be handled according to the specific circumstances, and this guideline offers ways that this could happen. I hope this answers your question. —Matrix(!) {user - talk? - _uselesscontributions} 17:11, 1 July 2024 (UTC)

I don't have strong feelings about this in any particular direction (which is one of the reasons I decided to close the DR). The numbers and weight of arguments in reference to policy seemed to mandate a keep. I also don't see a problem with Matrix's VP close. Is there some more precise aspect of it you wanted me to comment on? I caveat this by saying I'm not sure why my comment beyond the DR close should have any particular weight. IronGargoyle (talk) 17:35, 30 June 2024 (UTC)

@Rhododendrites, Matrix, and IronGargoyle: My question is how should this page be updated in light of the village pump discussion? Since it seems there was consensus for Proposal #4, should we modify the "How Commons handles copyleft trolls" to include those suggestions? In other words, should we replace the "No standard practice" section with a section saying that verified copyleft trolls should be asked to upgrade to CC4 and if they refuse their images should be watermarked? That would seem to be the logical conclusion, IMO. Nosferattus (talk) 00:21, 2 July 2024 (UTC)

That was largely my question above, too, but the reply from Matrix indicates, to me, that there is nothing prescriptive there. — Rhododendrites ^talk |  03:34, 2 July 2024 (UTC)

Well we seem to be caught in a catch-22. No one wants to take any action against copyleft trolls because there are no prescriptive guidelines. And there are no prescriptive guidelines because we aren't taking any consistent action against copyleft trolls. Proposal #4 seemed to have consensus, and that was confirmed by Matrix's closure. So shouldn't we add the wording from Proposal #4 to the "How Commons handles copyleft trolls" section? Otherwise, we are effectively enacting Proposal #5 (to just create a landing page), which did not have consensus. Nosferattus (talk) 21:37, 2 July 2024 (UTC)

@Matrix: Would you mind clarifying your comments? Nosferattus (talk) 21:39, 2 July 2024 (UTC)

There are a lot of implementation problems that still need to be discussed. For example, what's the threshold for a forced watermark? How do we put the watermark in bulk? What should it say? How will we deal with images still in use? I wanted to turn this into a guideline (since it still offers valuable information) but to turn this into anything prescriptive we still need details on how we would execute forced watermarks, we can't just say "put forced watermarks" and leave for the day. —Matrix(!) {user - talk? - _uselesscontributions} 14:54, 3 July 2024 (UTC)

Thank you! That is very helpful guidance. Nosferattus (talk) 03:11, 4 July 2024 (UTC)

What is the threshold for a forced watermark?

Per Commons:Village pump/Proposals#Proposal 4 and the discussion above, we need to decide what is the proper threshold for a forced watermark. My suggestion would be:

1. There is consensus that the author of the work(s) or their proxy is engaging in copyleft trolling per the definition given here. This must be established on a case-by-case basis through evidence and discussion.
2. Following the establishment of copyleft trolling (#1), the author of the work(s) does not agree to switch to a CC4 license within 30 days of being asked to do so. (If the author does not have an account on Commons, this can be asked on Flickr, or wherever the images were originally posted.)
3. If the author subsequently switches to a CC4 license, the watermarks can be removed.

Nosferattus (talk) 01:19, 7 July 2024 (UTC)

How do you intend to apply forced watermarks? Ok if there is a few images to add manually but if there is a few hundred or thousand is going to be impossible. Bidgee (talk) 08:57, 7 July 2024 (UTC)

ImageMagick. Nosferattus (talk) 20:02, 7 July 2024 (UTC)

I still don't get how CC4 is a solution here. Yes, for the time being it would mean you can't work with Pixsy but for every case where (1) is satisfied without using Pixsy, if they switch to CC4 they can just continue on as they have been. — Rhododendrites ^talk |  13:20, 7 July 2024 (UTC)

I assume that very few authors are willing to put in the time required to do this without Pixsy (manually search for misattributed images, hunt down the users, handle the negotiations, retain a lawyer if the case gets to court), so this only seems like a potentially sizeable issue if there are Pixsy competitors willing to work with CC4. Is that the case? I can try to find out if nobody has looked into this yet. -- Julesvernex2 (talk) 17:36, 7 July 2024 (UTC)

Philpot, one of the cases mentioned on this page, did not use Pixsy AFAIK. — Rhododendrites ^talk |  18:47, 7 July 2024 (UTC)

I don't know if CC4 is a good solution or not; it's just the only solution so far that has gotten any support. Please propose other solutions if you have ideas. Unfortunately, the Commons community seems to prefer shooting down ideas over finding solutions (and the perfect is always the enemy of the good). Nosferattus (talk) 20:00, 7 July 2024 (UTC)

I don't think CC4 is a solution. We should certainly encourage it across the board, and I would support a requirement that images created (and not just uploaded) after [whenever such a proposal would be enacted] cannot use older CC licenses across the project. But I don't think we're going to find a bright line for when forced watermarking is required because opinions vary so much about what should qualify as "copyleft trolling" and what the best approach to dealing with it is. We've thus far been very inconsistent. My hope is that having this guideline will provide a sort of rubric to evaluate future cases, and to establish that there are multiple possible remedies on the table. Many of the !votes in the Diliff DR talked about the practice not going against any Commons policy or guideline. Well, that's different now, to an extent. Not arguing to relitigate that case, but it'll be interesting to see if the next case will be different.

Here's a constructive next step you could take: establish, for the cases when we do use watermarking to protect users against copyleft trolling (regardless of how those cases are determined), how exactly should this watermarking take place. Maybe you could work on an essay/information page on the subject, trying to sort out all the practical bits (wording, placement, coloring, size, tips for adding them, information on how Wikipedias can crop using templates, etc.). I wouldn't take the Philpot case as prescriptive, but an example of how it's been done before. That text never really went through a proper discussion AFAIK, so whatever the discussions that happen over this hypothetical information page find may result in those watermarks being changed to align with TBD best practices. — Rhododendrites ^talk |  20:16, 7 July 2024 (UTC)

@Rhododendrites: Some other ideas for condition #2:
A. Following the establishment of copyleft trolling (#1), the author of the work(s) does not agree to allow good faith reusers a 14-day grace period to correct attribution errors before demanding a fee or initiating legal action.
B. Following the establishment of copyleft trolling (#1), the author of the work(s) does not agree to discontinue the problematic actions identified in the discussion.
Do either of those sound more workable? Nosferattus (talk) 20:28, 7 July 2024 (UTC)

Those suggestions sound workable to me, but I would suggest to begin one step beforehand: How are cases of copyleft trolling identified in the first place?

1. An original uploader (following OUL) can be reported for copyleft trolling. Those reports (by one or by multiple named re-users, irrespective of having an account on Commons) need to be credible and substantiated, ideally by handing in the evidence/emails/legal documents to the VRT who then acknowledge the established legal threat in the discussion thread. (Nothing speaks against showing the whole legal correspondence in public, but VRT seems to be the best way to ensure the privacy of the reporting re-user (following RRU), if that is wanted.)
2. A discussion on the case is then held on the talk page, ideally with participation of both the OUL and RRU. This discussion has to take at least a certain time - 2 or 4 weeks. Possible results include the OUL (or the VRT) disproving the allegations of the RRU (false reports/misunderstandings?); the OUL agreeing to no longer engage in copyleft trolling (by switching to CC4, or also informally); or the OUL insisting to pursue his copyleft rights (--> establishment) or the OUL refusing to participate at all (--> establishment). Nothing in our process here, changes anything about the civil legal proceedings and obligations of the RRU who got sued.
   • (I don't think anyone will argue that the case examples of Philpot or Diliff warrant us to go through steps #1 and #2 again: These are established cases of copyleft trolling.)
3. Once it is established that a user has engaged in copyleft trolling per steps #1 and #2 and seems likely to continue that practice, actions can get taken, possibly including Deletion Requests or Forced Watermarking.

With regards for subsequent removal of the forced watermarks if something changes... I am doubtful that a switch to CC-4 really cures the problems of copyleft trolling. So a mere license switch should not be grounds for watermark removal, the OUL needs to also promise to not continue trolling. --Enyavar (talk) 10:55, 19 July 2024 (UTC)

@Enyavar: I think your suggestions about the establishment process are fine, although I don't think we have to be that detailed in prescribing the process. Basically, there just needs to be consensus that the user is copyleft trolling and the user must be given an opportunity to defend or explain their actions. Nosferattus (talk) 19:23, 3 August 2024 (UTC)

@Rhododendrites, Julesvernex2, Enyavar, and Chris.sherlock2: From the discussions here so far, it sounds like there is agreement that switching to the CC4 license is not a solution to copyleft trolling. How do people feel about requiring agreement to a 14-day grace period instead (and otherwise applying forced watermarks)? Nosferattus (talk) 16:46, 5 August 2024 (UTC)

a misunderstanding that belongs to a different policy, has been moved to a new section below

We just can't prevent copyleft trolling. All we can do here is collect evidence about who does it, try to friendly convince them to stop it, and if they won't, we delete or watermark their uploads. --Enyavar (talk) 18:42, 5 August 2024 (UTC)

I'm not suggesting that Commons try to enforce a 14-day grace period on all uploads. I'm suggesting that after a person has been found to be copyleft trolling, if they don't agree to allow a 14-day grace period from then on, then we add forced watermarks. This is basically proposal A from above. It sounds like maybe you would prefer proposal B. Could you clarify? As far as getting the WMF to change their ToU (similar to what Flickr did), that was already proposed on the Village Pump and shot down by pretty much everyone except Glrx, so I don't think it will fly. Nosferattus (talk) 15:25, 8 August 2024 (UTC)

Yes, it seems we may have had a misunderstanding there. I now moved that part into a new section (see here), and also move the discussion about how the watermarks are going to be formed (see here). --Enyavar (talk) 15:34, 8 August 2024 (UTC)

Okay, now that I removed the distractions to another section, here's my reply: Yes, I'm with proposal (b). Because (a) means to unilaterally change the Terms of Service for just the offending individual users, like Iliff. That would be unenforceable and/or possibly even unlawful, see the moved section.

Rather, we could give the copyleft troll a grace period of 14 days after our (b)-consensus is reached. If the troll uploaders have not credibly assured within in 14 days that their affiliation with Pixsy is over, the community may proceed to pursue watermarking or deletion. (note: may, not must. In the case of Diliff, we are late by months; and I wouldn't expect that future cases are clear-cut.) --Enyavar (talk) 18:21, 8 August 2024 (UTC)

Can Commons deal with copyleft trolls by changing the Terms of Use?

This section was created after a misunderstanding.

There is no legal framework for us to require such agreements [i.e. 14-day grace periods]. The CC4 license prescribes a whole month grace period. The CC3 license (and earlier ones) prescribes no grace period at all, which means that Pixsy may legally serve you a fee the next second after they detected your using a CC1-CC3 picture with improper attribution. How do you imagine we as a platform would approach Pixsy to grant detected re-users two weeks, when that very idea runs both counter to their business model and counter to the license model they misuse? Who are even the negotiating parties in such a deal? WMF and Pixsy? People like Iliff who use Pixsy? The re-users? The good lawyers from Creative Commons? Compared to all of them, we concerned users on this talk page have about the least binding power to decide any grace periods for licenses. --Enyavar (talk) 18:42, 5 August 2024 (UTC)

Yes, there is a legal framework. WMF can set the terms of use of its websites. If you upload a file here after the new terms take effect, then you agree to give any copyright infringer 30 days (not 14) to correct the error. Glrx (talk) 19:26, 5 August 2024 (UTC)

And would WMF possibly enforce those ToU? The action of all copyleft trolling happens completely outside of Commons, we are just hosting the bait. Also, what happens to files our users uploaded before the date of the changed ToU? --Enyavar (talk) 12:45, 7 August 2024 (UTC)

WMF would not need to enforce that portion of the ToU. It is a contract with a third-party benefit. The defendant would show the clause applies and claim its benefit. It is not a complete solution, but the ToU can be if you upload a file after date x, then you agree to the ToU for even earlier uploads. Glrx (talk) 18:42, 7 August 2024 (UTC)

@Enyavar: I think you misunderstood my suggestion. I'm not suggesting that Commons try to enforce a 14-day grace period on all uploads. [..] As far as getting the WMF to change their ToU (similar to what Flickr did), that was already proposed on the Village Pump and shot down by pretty much everyone except Glrx, so I don't think it will fly. Nosferattus (talk) 15:25, 8 August 2024 (UTC)

Sorry for the misunderstanding. Yeah, I don't think that a general change in the ToU (Terms of Use; or Service, ToS) would work all that great in practice, because Commons/WMF would need to roll out notifications to everyone. Then, would Pixsy and their users even follow those new restrictions? Also, what about the already uploaded pictures? We would have wait until this is litigated by the courts - for images pre-ToU-change and post-ToU-change, for each of the CC licenses, and for various other circumstances (personal use vs. private use, etc.). Until such litigation is done, Pixsy (and their contractors) can still rake in the money they extract from users who are afraid to litigate the issue, and may even be rightfully afraid.

But if we demand just from a few uploaders that they accept a unilateral change in their ToS (requiring them and their contractor, Pixsy, to observe new rules made up by the Commons community that are not grounded in any license or law), such a policy means unequal treatment due to off-platform behaviour. I am not a lawyer, but if Pixsy takes the requirement to observe new grace periods seriously, then I anticipate them to formulate a discrimination lawsuit against WMF or something similar. Or, if they don't take it seriously, they will make uploaders like Iliff agree to the requirement, and then just ignore that unenforceable agreement. Or, they will just ignore us entirely. --Enyavar (talk) 18:21, 8 August 2024 (UTC)

It doesn't matter what Pixsy thinks of the agreement. If the uploader agrees, that almost certainly means they will need to discontinue using Pixsy (unless they can somehow get Pixsy to ask people to correct their licensing before threatening them with lawsuits, which seems unlikely). The rules on Commons don't have to be grounded in a license or law. We can make up whatever rules we want as long as it doesn't break existing laws. Nosferattus (talk) 20:44, 9 August 2024 (UTC)

How should the watermarks look?

If watermarks are affixed to the images, they must include the entire required attribution. To discourage the practice, a warning should also be included. Okay, now how should watermarks look like? I'm not sure that the Philpot images are establishing a good practice: The attribution texts in those are very large. I uploaded an attribution test example on a currently not-in-use image by Diliff for demonstration purposes, it includes a) the desired attribution in bold and b) a warning. Technically, I added a 94 pixel white border below the proper image with black text, Verdana, Font-size 36. Added is the important word, because I did not change a single pixel of the original image. The chosen text of the warning stands up for debate, too. The following would have been like Philpot: "David Iliff sued users of his work for minor attribution errors. Keep the attribution intact." As far as I know, Iliff did not sue however, he threatened lawsuits in the legal notices; and it also has not been proven that the attribution errors were in fact "minor" (afaik: we just know there were errors). That's why I rephrased it as "Caution for re-users: Keep this attribution intact, Iliff has been threatening users of his work with lawsuits for infringement." I'm open for better suggestion... for example "copyright infringement" sounds better? Regarding the size: this is a in a very large font already but proportionally to the image it seems rather small. With the smallest download size (301 × 240 pixels), it can only just be read if you know what it is supposed to say. That might not suffice; but on the other hand, re-users are still responsible in how they re-use images.

@Enyavar: I think the small size of the watermark is OK. It just needs to be there, it doesn't need to be super prominent. I also agree that it needs to include the attribution, the license, and a brief warning to reusers to retain the attribution. I think saying "copyright infringement" rather than just "infringement" is important, as "infringement" is too vague about what is going on. Nosferattus (talk) 19:23, 3 August 2024 (UTC)

Do we have a consensus on the final wording for the watermark on Iliff's 1000+ bait images, before there is more harm? Is "Caution for re-users: Keep this attribution intact, Iliff threatened users of his work with lawsuits for copyright infringement." acceptable? And how will we apply the standardized watermark on all his images? Let's be real, Iliffs "grace period" here on Commons is about to approach 6 months. --Enyavar (talk) 18:42, 5 August 2024 (UTC)

Reply to Enyavar's earlier question: Applying the watermark to all of Diliff's images could be done with ImageMagick and a bot. Nosferattus (talk) 15:30, 8 August 2024 (UTC)

You mentioned that program before, I trust you have a fitting script ready? Some pictures are larger than others, so the added whitespace and watermark may need to be scaled differently. --Enyavar (talk) 18:21, 8 August 2024 (UTC)

The images are still up, and more people have stumbled into this deliberate trap. @Nosferattus: , when are you going to start up the ImageMagick bot and process Iliffs images? --Enyavar (talk) 22:52, 19 March 2025 (UTC)

@Enyavar: I just added an example ImageMagick script to Commons:Copyleft trolling#Forced watermarking. Should I create a bot request to run it on Diliff's images? Nosferattus (talk) 05:15, 20 March 2025 (UTC)

Could we first see the results in an example or two? The script you linked looks okay to achieve what we want, but I'd like to be cautious. --Enyavar (talk) 05:52, 20 March 2025 (UTC)

@Enyavar: Here is an example of what the script outputs. Nosferattus (talk) 06:05, 20 March 2025 (UTC)

When scaling down, yours is better legible than the one I produced with gimp, see above; but they're both barely visible when thumbnailing (some wikis prefer standard thumbnails of 200px images). Images that are in used in wikis with such thumbnails, would have added "chickenscratch" notices, but I think that is fine: less obtrusive in wiki usage; but still noticable once you download the images for own usage. "Leave attribution intact" is straightforward, but potential reusers might not realize the reason. Can we include a shortcut to the Copyleft Trolling page? "Commons:CLTROLL" or "Commons:CLTR"? --Enyavar (talk) 08:28, 20 March 2025 (UTC)

There are a few problems with referring to Commons in the watermark and I don't think it's really needed anyway. First, I can't create an actual link, and no one besides a Wikimedian is going to understand what "Commons:CLTROLL" means. Also, the more text we put in the watermark, the smaller I have to make it to ensure it will fit. (Fitting the text involves a fair bit of guesswork, as ImageMagick is a very bare-bones tool, unlike Gimp or Photoshop.) Also, I'm trying to keep the watermark as apolitical (in the Wikimedian sense) as possible, so we can build consensus around using it. The important thing is that the reuser knows how to properly attribute the image so they don't get extorted by the copyleft troll. I think this accomplishes that (even if they don't understand why it's important). Nosferattus (talk) 15:32, 20 March 2025 (UTC)

Just a reminder there was no consensus to apply forced watermarking to Diliff's images. — Rhododendrites ^talk |  12:31, 20 March 2025 (UTC)

@Rhododendrites: Thanks for the reminder. I guess we still have a lot of work to do. First of which is deciding on an appropriate watermark. Do you have any opinion on the ones presented here? Nosferattus (talk) 15:35, 20 March 2025 (UTC)

On style, I don't have a strong preference. But it's worth noting that a one-size-fits-all solution may be difficult. My understanding is that the strict interpretation of the license requires either a full copy of the license of a link/url. I don't think many people actually care about that, but it's a potential gotcha. Diliff's instructions actually explicitly require doing so, but then that part is omitted from his "suggested attribution" line, so it's unclear. Also just providing the name of the photographer sometimes isn't enough. e.g. with Philpot and Verch there was more required (full urls). Finally, there should be a line in there about why the watermark is included -- one that's flexible enough to be included anywhere. How about "This watermark was added by Wikimedia Commons following multiple allegations that the photographer used threat of legal action to extract payment from media users for inadequate attribution. If you crop this watermark, attribute exactly as presented here." — Rhododendrites ^talk |  15:48, 20 March 2025 (UTC)

Good points. I'm not sure what to do about the link requirement, but most people don't follow it anyway, and I doubt Pixsy cares about that aspect. I like your suggestion about the warning text. Unfortunately, I don't know how to automate appending such a long block of text (that would need to wrap) to an image of arbitrary size. Is there anyone else who could help with that? Nosferattus (talk) 16:08, 20 March 2025 (UTC)

BTW I removed the script from the page -- not because there's any problem with sharing it, but because scripts on guideline pages is just weird and guidelines need to be as accessible to readers as possible. If you want to put instructions on a userspace subpage or something we could link to it in see also, I suppose. — Rhododendrites ^talk |  12:35, 20 March 2025 (UTC)

I moved the script to a subpage. I didn't put it in userspace because I don't want to own it and it could definitely be improved with further collaboration. Nosferattus (talk) 15:22, 20 March 2025 (UTC)

there was no consensus to apply forced watermarking to Diliff's images - afair, there was no consensus to have them all deleted. But what we have here is a clear case of continued copyleft trolling, and we had a consensus that we don't want to encourage this behaviour, too. After all, this page was created based on Diliff's case with an 11-1-1 voting that supported it. That seems a lot like consensus to me.
That said, Iliff himself suggests the exact text that Nosferattus could place in the bird picture above. I do think that would be enough, given how longer texts create problems for the convenient script. Maybe Nosferattus can test how "Keep this attribution intact to avoid legal action." works. The license url itself, is quite short as well. Since Iliff seems to require it, there could be a second line added below each image:

• https://creativecommons.org/licenses/by/3.0/ (for most of Iliffs uploads)
• https://creativecommons.org/licenses/by-sa/4.0/ (for images like File:South Bay Beach, Hong Kong - Diliff.jpg, I am not kidding, there are cases where Iliff published the image under 4.0 while still also requiring the regular attribution text for 3.0)

That the links do not work until you retype them or OCR them, does not make them invalid, and they are short enough to fit in one line. --Enyavar (talk) 09:02, 22 March 2025 (UTC)

There didn't seem to be consensus for anything other than creating this page. But this brings something up: it might be useful to have an organized list of discussions relevant to this page (big discussions on the village pump, but also deletion requests and even messages from media reusers to uploaders like the diff above). That might make it easier to highlight a pattern in the future, should it come to that. I'm not talking about a list of diffs on the guidelines page itself, but potentially a subpage or a dedicated talk page section. — Rhododendrites ^talk |  15:01, 22 March 2025 (UTC)

@Rhododendrites: Actually looking back at the discussion, it looks like there was clear consensus to apply forced watermarks to authors who refused to switch to CC-4.0 licenses (which is the case here). Looking even further back, there was also consensus to ban uploads from clients of Pixsy, although that discussion was never officially closed for some reason. We have spent countless hours debating and wringing our hands, but haven't done a single thing to actually limit DIliff's copyleft trolling. This seems like a classic case of Wikimedians refusing to enforce rules on vested contributors at the expense of the project. What is your advice on how to move this process forward? Nosferattus (talk) 15:21, 23 March 2025 (UTC)

Ah yes I remember. The proposal was based on a misunderstanding of CC4 and got a bunch of support before this was clarified. To reiterate: switching to CC4 does not provide a 30 day grace period to fix an error before money can be demanded. It doesn't prevent anyone from being sued in exactly the same way as previous CC versions. That wasn't corrected until the end of the discussion, after which point there was no more support. It merely provides a window to restore the license after a violation. i.e. if I misuse your photo, you demand money, I pay money, and I fix attribution, under CC3 you can sue me again because the license was never restored while under CC4 I have a window in which to "cure" it. It only prevents me from being sued again, in other words. I don't recall if I asked Matrix to revise their close at the time or if I just didn't see it (more than a month after discussion ended), but the guideline reflects reality. Absent CC4 doing the thing people supporting that measure thought it did, it doesn't make sense to view that proposal (#4) as anything other than support for a guideline (which we have now) and support for a fix that doesn't actually exist.
That thread about banning Pixsy is interesting; I've never seen that before (another good reason why we need a centralized index of these discussions). It doesn't look like it was closed, and I'm not sure how to interpret it. Doesn't look like we actually followed through and banned anything, right? How would that even work? Someone shares a demand letter and we delete the file? Ban users represented by Pixsy? Watermark the images and ban the user? None of that is clear and needs to be resolved somewhere with more watchers than this page. — Rhododendrites ^talk |  15:49, 23 March 2025 (UTC)

@Rhododendrites: I  Support "Watermark the images and ban the user".   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 16:17, 23 March 2025 (UTC)

Which images and users? Anyone who uses Pixsy? To better understand how Pixsy works before the above talk, I created an account and uploaded one image. Do I get banned? What if used it to send exactly one demand letter to AOL when it didn't credit me? (To be clear, I'm not going to be using Pixsy for anything.) Does it also require consensus that the user is engaged in copyleft trolling? What are the processes and thresholds? But again, this seems like something that needs to happen at a village pump (even if to formalize/operationalize the outcome of the previous discussion). — Rhododendrites ^talk |  16:28, 23 March 2025 (UTC)

@Rhododendrites: I don't entirely agree with your interpretation. Regardless of the actual remedies available under CC4, it is still the case that Pixsy doesn't make legal threats against CC4 reusers. That fact alone makes a significant difference as most people are not going to bother tracking down reusers themselves. This is also the reason that Diliff refuses to switch licenses. The request to switch to CC4 is thus not entirely without merit and I don't think there's any reason to dismiss the previous consensus around it, at least until a better consensus is established. Nosferattus (talk) 16:23, 23 March 2025 (UTC)

This much is true. Nonetheless, it doesn't do what it was presented as doing in that discussion. — Rhododendrites ^talk |  16:28, 23 March 2025 (UTC)

Ironically, this was at least partially due to information presented on Pixsy itself: "However, users of images with a Creative Commons license release type 4.0 (or above) are granted a 30-day grace period in which they can resolve breaches of license terms, before it is considered copyright infringement." Nosferattus (talk) 17:19, 23 March 2025 (UTC)

Yes, and in my emails with Pixsy they said much the same thing. I did not correct them, of course, but I wonder if they'll wind up changing their policy once they realize it doesn't do that. :/ — Rhododendrites ^talk |  17:50, 23 March 2025 (UTC)

@Enyavar: Would a shorter link to COM:L like https://w.wiki/bk7 suffice?   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 10:09, 22 March 2025 (UTC)

Nosferattus already said There are a few problems with referring to Commons in the watermark, and after rethinking it, I agree. Once an image is reused on third-party sites, it makes no sense to refer to Commons policies. A direct link to the CC license however is apparently both required and feasible without shortening. I also thought about including a watermark via QR code, but that's an additional hurdle rather than a human-readable solution. --11:50, 22 March 2025 (UTC) Enyavar (talk) 11:50, 22 March 2025 (UTC)

@Enyavar and Rhododendrites: I've updated the caption script based on your feedback. An example of the output can be seen in the bird image above (which has been refreshed). Nosferattus (talk) 16:15, 23 March 2025 (UTC)

@Nosferattus: Your current watermark includes "Photo by DAVID ILLIF. License: CC-BY-SA 3.0. Keep this attribution intact to avoid legal action." and "License terms: https://creativecommons.org/licenses/by/3.0/". The license for that photo is actually CC-BY 3.0, not CC-BY-SA 3.0, but that is our mumbo-jumbo which does not need to be in the watermark. If you can fit "Photo by DAVID ILLIF. Keep this attribution intact to avoid legal action. License: https://creativecommons.org/licenses/by/3.0/" on one line, why not just use that?   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 17:36, 23 March 2025 (UTC)

@Jeff G.: Thanks for spotting that! Just goes to show how easy it is to make licensing mistakes. Since the license only technically requires providing the URL to the license, I've followed your advice and used the URL instead of the license short name. I'm still breaking it into 2 separate lines just in case the attribution string needs to be longer for future cases. This also makes it easier to crop the warning off if you want to keep the attribution and license but not the warning. Nosferattus (talk) 19:15, 23 March 2025 (UTC)

Thanks, I think this looks okay now. Are we participants on this talk page all agreed that this is the watermarking variant that we are going to present on the village pump, before finally watermarking Iliff's photos? --Enyavar (talk) 20:36, 23 March 2025 (UTC)

 Support.   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 21:16, 23 March 2025 (UTC)

 Done here: Commons:Village_pump/Copyright#Copyleft trolling - proceeding to watermark images. If we don't get much attention there, we could also link to that place from the Proposals section, or from the main pump. --Enyavar (talk) 16:03, 26 March 2025 (UTC)

discussion index

Started here: Commons:Copyleft trolling/discussion index. Please add. Let's not link from the guideline, yet, though. — Rhododendrites ^talk |  01:15, 28 March 2025 (UTC)

Name of the policy page

The page title of "trolling" implies a maliciousness that is not necessarily present when photographers are (maybe?) honestly trying to protect the work which they uploaded. Uploader like Diliff believe their work is usally stolen by commercial moneygrabbers, and they'd be offended when called moneygrabbers themselves.

So I think that we're not really doing ourself a service by calling these enforcement practices "trolling". Maybe we could call it "Copyleft enforcement". Many people have agreed previously that the people who try to "enforce copyleft" have the full legal right to do so (and I would also agree) - we just were of different minds whether Commons should allow the practice, or not.

We would now begin to watermark pictures where users have been shown to engage in "Copyleft Enforcement", because we are trying to protect re-users, which is in turn also the full right of our community that endorses media freedom. We should state that we are not explicitly going to punish the users who do it, except maybe in extreme cases. We just state that we need to enforce the right of re-users to get fair warning before accidentally using photos which were made by hardened "Copyleft Enforcers". Again, the watermarks are not serving as a punishment of the creators, but a safety mechanism for everyone else.

So, can we rename the page, and partly rewrite it with this concept in mind? --Enyavar (talk) 16:55, 5 April 2025 (UTC)

 Support.   — 🇺🇦Jeff G. ツ please ping or talk to me🇺🇦 17:05, 5 April 2025 (UTC)

 Support I never liked the phrasing myself. "Copyleft Enforcers" is good because it keeps the person's motivations out of the conversation though. --Adamant1 (talk) 18:56, 5 April 2025 (UTC)

•  Oppose This feels like an extreme move. IMO this needs more eyeballs. See Commons:Village_pump#Copyleft_enforcement_-_concern_about_stretching_of_a_guideline. — Rhododendrites ^talk |  20:35, 5 April 2025 (UTC)

 Oppose If we bar enforcing the license, why are we using the license? We may as well ban anything that is not CC0, if all of our uploads come with the asterisk that actually the original creator has no way to enforce their rights. CC is not public domain. PARAKANYAA (talk) 20:42, 5 April 2025 (UTC)

 Comment It's not ANY form of Copyleft Enforcement that our guideline is intended to act against. The page lines out that there is a grey zone, and that we have to carefully consider each case. But we SHOULD act against those forms that directly ask for large lump sums of money due to minor licencing errors based on outdated CC licenses. Just as our guideline states on the reverse page. But we can still act against actual trolling while still calling it by a "nicer" name. @PARAKANYAA: I think you might have come upon this without prior knowledge of the cases and the guideline? We are NOT barring license enforcement. We try to stop the excesses of the Copyleft Trolling industry, and my suggestion was just to STOP calling every suspect a troll right from the start, and not treat them like villains. Nobody deserves that treatment... just as nobody deserves to get a bill out of the blue, over 800-1500 ₤$€ for minor mistakes, too. --Enyavar (talk) 21:15, 5 April 2025 (UTC)

Then why would we rename it, indicating that all forms of enforcement are the problem? The way this is written would indicate it so. PARAKANYAA (talk) 21:22, 5 April 2025 (UTC)

We would rename it to stop the immediate suspicion that anyone enforcing a license is necessarily a "troll" (i.e. the villain of the story). The grey zone exists. There are outright trolls; there are "law" agencies sending mass e-mails; there are specialized (actual) lawyers; there are those creators who first send a polite letter before the big bills. The last one is the option we'd prefer here. This policy page establishes (or is intended to do so) which of these practices would result in which treatment of your images, given that Commons is an open media archive. A page title that outright says "you're the bad guy if you enforce..." is just not helpful. And my intent with this proposal was to reflect that. --Enyavar (talk) 21:41, 5 April 2025 (UTC)

Appreciate this, Enyavar. As I said over at VP, I appreciate this sentiment, but it does seem like this would at minimum require a rewrite to fit the title, which is a pretty serious change, and I just don't think consensus is clear on that. It's clear there's consensus to do something about copyleft trolling, and it's less clear where to draw the line or what to do in edge cases. I think perhaps if the intention is just to be clearer that not everybody that falls outside of community expectations is necessarily "trolling", we could probably just use the section on "Copyright enforcement vs. copyleft trolling" to articulate that better. From your perspective, I could see where my VP thread took this wrong, and I apologize for that, but this thread following some of the comments here and at VPC gave me bad feelings that process was getting lost in pursuit of a remedy, and IMO process is important with these complex cases. FWIW. — Rhododendrites ^talk |  21:52, 5 April 2025 (UTC)

•  Strong oppose: your heart is clearly in the right place but I strongly disagree. Copyleft trolling and copyleft enforcement are complete opposites. If we'd phrase it in terms of enforcement, copyleft trolling would be copyright enforcement following a copyleft trap. We actually want and should be incentivizing everyone to practice copyleft enforcement, which is quite in line with CC 4.0's terms that state that the license is reinstated if the transgressor complies with the terms of the license after learning about his transgression. We could very well have a voluntary team on copyleft enforcement to notify people misusing CC licenses and instruct them on how to fix it. If they deliberately don't, they're simply practicing copyright violation pure and simple, and the owner has every right to pursue compensation however they see fit. Even this would be in accordance with our mission and beneficial to the project; none of this is trolling, at all. Trolls don't want to enforce anything, they want to set up traps and profit from it. That's why they still favor CC 3.0, that's why they pursue even minor attribution mistakes with no prior communication, that's why they ignore talk page messages by people honestly trying to get it right. To sum up: we're against copyleft trolling but absolutely for copyleft enforcement. The grey zone is much narrower than you portray it, and users should be given the benefit of the doubt. Legitimately enforcing copyright after deliberate and reiterated transgressions of the CC license is absolutely acceptable - desirable, even. Rkieferbaum (talk) 16:44, 7 April 2025 (UTC)
•  Oppose - Copyleft enforcement isn't a problem. Copyleft trolling is. Renaming the page would muddy the waters, IMO. If Diliff (or anyone else) doesn't want to be called a troll, they can just stop copyleft trolling. Nosferattus (talk) 16:57, 7 April 2025 (UTC)

Defining accepted copyright enforcement

I would propose that we define copyright enforcement under the following conditions as not being copyleft trolling, but all that goes beyond this as copyleft trolling.

Demanding money for violations of the license terms is only allowed in the following cases:

1. If the re user got 14 days to fix the attribution but did not so. May be extended to 30 days if the re user requests a deadline extension.
2. In case of a print if they do not fix the attribution in a new print run after being informed of the license violation. Destroying of existing prints must not be demanded. This also applies to DVDs or similar.
3. Immediately if the violation is made by a stock photo company.
4. Immediately if the violation is made in an advertisement of a large company.
5. Immediately in cases of copyfraud if clearly not accidentally.

GPSLeo (talk) 07:19, 8 April 2025 (UTC)

 Support the definition in general. There are a few words I'm not fully in accordance with.

• "is only allowed" implies that acting counter to the guideline is forbidden on Commons. Since we're not writing a legal text, "okay" instead of "allowed" seems better.
• (#1) "14 days after notification"
• (#2) "print run after notification. Destruction of"
• (#5) what are "clearly not accidental cases of copyfraud" ? This again leaves room for interpretation. If I take a photo from Commons without attribution, resize it, add a meme and post it on Social Media, for the purpose of amassing 20 or maybe even 2000 likes? I'd argue that is a clear case of not being media-literate, but "copyfraud"? Pixsy is known to go have gone after social media posts in the past. I'd be in favor to strike case 5, or maybe replace it with "Immediately in cases where the re-user can be shown to have directly gained from the violation, at least equivalent or more than the money demanded." Maybe different wording?
• Afterwards: "If users provably enforce their copyright in ways that are not okay (as per the listed exceptions above), their enforcement behaviour is considered "copyleft trolling". Depending on the extent of the behaviour, there are various options for the community. [etc: verification of user complaints, dissuasion, watermarking. Extreme cases go further.]"

--Enyavar (talk) 14:50, 8 April 2025 (UTC)

Our guidelines are kind of legal texts, if users do not follow them they are banned from the project. With clear cases of copyfraud I mean that someone says that they own the copyright or even re-license the work. Posting a photo on social media without giving any source is a simple license violation but not copyfraud. If the post says "Look at the photo I made" it would be copyfraud. GPSLeo (talk) 15:48, 8 April 2025 (UTC)

Ah, I understand. Your usage of "copyfraud" is clear to me now, but maybe not to everyone else reading the text later.

On the topic of bans: No, I don't think users engaging in Copyright trolling should be banned. That should still be the last remedy for extreme cases. Take Diliff: he is in the grey zone where we would not punish him with a ban, but watermark photos regardless. --Enyavar (talk) 15:59, 8 April 2025 (UTC)

This is way too complicated. All we need to say is "Uploaders must allow good faith reusers the opportunity to correct errors." Stock photo companies and advertising agencies have lawyers and understand licensing, so if they misuse a Commons image they are not acting in good faith. Nosferattus (talk) 22:48, 8 April 2025 (UTC)

2026 update: Pixsy still pursuing Wikimedia CC-BY images with IPEC threats (UK)

2026 update: Pixsy still pursuing Wikimedia CC-BY images with IPEC threats (UK). No reported judgments despite multiple demands. Can we clarify policy on photographers who systematically bait attribution violations then pursue settlements? Ppp123qqq (talk) 13:32, 4 March 2026 (UTC)

To clarify, the current policy is "we give warm words for the victims and free reign for the trolls."

Yes, the page reads differently, but the community has not been willing to do anything for the last two years, in order to not lose "valued photographers", not even to protect the integrity of the project. Anyone who wants to make a big buck, can upload copyleft content and sue unsuspecting re-users. That's it.

Unless others were active in the background, we're also completely blind. We have no action policy for found-out trolls (although the two publicly uncovered ones were seemingly harrassed off the platform), and we don't even keep lists of suspected and/or confirmed cases. The community failed to agree on any systematic action, which means the trolls are given free reign. I have no doubt that they weere not idle.

I'm still willing to help change that, but not with zero support. --Enyavar (talk) 23:44, 4 March 2026 (UTC)

Community wants progress @Enyavar.

Tupungato: Strong first candidate for your troll list

== Friend reports Pixsy £900 demand against Tupungato Commons image (UK 2026) ==

User:Tupungato (Special:Contributions/Tupungato): 1,100+ CC BY 2.0/3.0 files.

Reported pattern:

• Friend received Pixsy £900 demand for "attribution" despite credits page + prompt removal
• General Pixsy/Wikimedia claims: [Reddit £500 CC demand] | [LegalBeagles IPEC threats]

Matches Commons:Copyleft_trolling exactly.

Support offered:

• Public troll list starting with Tupungato
• Pixsy report template (privacy protected)
• Flickr banned Pixsy trolls
• CC2.0/3.0 warning notices

Requested:

• Admin review → file deprecation/watermarking
• Licence upgrade to CC4.0

Private evidence via VRT.

--Pomme7755 (talk) 01:02, 5 March 2026 (UTC)

The most important information is missing here: What is the context the file was used? Was it used on a personal blog or by a commercial company? And what was the problem with the attribute that led to the demand? GPSLeo (talk) 19:17, 5 March 2026 (UTC)

@GPSLeo: The context is a small business website publishing a general informational blog post. An intern selected the image from what they understood to be a free Creative Commons source on Wikimedia Commons, as a generic illustration among others – it was not central to the article, more like a side comment. All third‑party sources were listed on a general credits page (including Wikimedia Commons). When the Pixsy demand arrived, both the article and image were removed promptly.

So:

• Yes, small business context – but the file was offered under a free CC licence on Commons.
• Attribution existed (via a credits page) and was easily fixable had a warning been given.
• No prior notice or chance to correct was offered – only a £900 demand and IPEC threats.

The concern is uploaders who use Commons' free CC platform, then immediately route minor attribution issues to Pixsy for non‑negotiable cash demands (£900 here), rather than contacting re‑users directly to request compliance.

That "upload free → Pixsy demand → no grace period" pattern matches Commons:Copyleft_trolling exactly. A troll list and CC2.0/3.0 warnings would prevent this.

--Pomme7755 (talk) 21:18, 5 March 2026 (UTC)

Was there a link from the photo to the credit page as an html link or something like a (1) in the caption? GPSLeo (talk) 05:04, 6 March 2026 (UTC)

@GPSLeo: Friend's small company blog post. Generic CC image from Commons (not central), credited on general credits page, removed on Pixsy demand. No inline "(1)" link, but fixable with notice.

If attribution was the goal, first step is email requesting correction. Instead: £900 Pixsy demand + IPEC threats, no grace period.

Honest uploaders request fixes. Pixsy cash demands for fixable CC attribution = trolling.

Reputational risk: When re‑users are hit with large Pixsy demands over fixable attribution issues on “free” Commons images, they don’t distinguish between an individual uploader and the platform – they just decide Wikimedia/Wikipedia are risky to use or link to vs paid stock sites (which at least disclose costs upfront).

Proven fixes:

• Flickr banned major Pixsy troll (27k images)

Commons solutions exist:

• File deprecation
• Watermarking
• Licence upgrade (CC2.0/3.0 → CC4.0)
• Upload blocks

Community needs to use them vs Pixsy‑pattern uploaders.

Troll list identifies patterns. Community action protects Commons' reputation.

--Pomme7755 (talk) 09:54, 6 March 2026 (UTC)

“credited on general credits page” Attribution must be on the same page, not a separate page.

this is what is stated in the legal code for CC-BY 2.0

”at a minimum such credit will appear where any other comparable authorship credit appears and in a manner at least as prominent as such other comparable authorship credit.”

typically this is within the caption of the photograph/image being used or in some unconventional manner is on the same page but prominently displayed. Bidgee (talk) 17:02, 6 March 2026 (UTC)

It need not be made on the same page. The license allows the credit to be made on a separate page if that is how other contributions are credited. The apparent issue here is the work was credited to Wikimedia Commons rather than the copyright owner. Glrx (talk) 17:13, 6 March 2026 (UTC)

Thanks both – the exact placement / wording of the credit is useful to clarify, but that’s precisely why this feels like a copyleft trolling problem rather than a normal licence issue. This was a fixable attribution mistake on a “free” CC image: there was a credits page, the use was non‑central, and the content was removed immediately once the problem was raised.

There was never any attempt by the uploader (or Pixsy) to ask for a correction or to point out what they wanted the credit to look like. The very first and only contact was a £900, non‑negotiable demand with IPEC threats.

Whether the credit should have been on the same page or the credits page is important for licence compliance, but the trolling concern is about the pattern: free CC uploads → no grace period → large Pixsy cash demands for fixable attribution issues.

That is what needs a Commons response (troll list, deprecation/warnings), alongside educating re‑users about best‑practice attribution. Pomme7755 (talk) 18:29, 6 March 2026 (UTC)

@Bidgee @Glrx: Thanks – attribution placement useful. Credits page vs inline common small-site practice Pixsy treats as £900 violation.

Core trolling per guideline: "refusing to allow violators to fix errors rather than pay a fee".^[1]

Here: free CC → no correction request → £900 Pixsy/IPEC → no grace despite prompt removal. We must protect Commons integrity from becoming a money-making tool for unscrupulous uploaders.

@Enyavar @Nosferattus @Jeff G. @Rhododendrites: Your great ideas on troll lists/watermarking/Flickr shouldn't stay just ideas—they need action now. Tupungato perfect test case.

Let's action it:

• Add User:Tupungato to troll list
• Deprecate Category:Images by Tupungato
• Private VRT evidence available

--Pomme7755 (talk) 21:20, 6 March 2026 (UTC)

Pomme7755 (talk) 21:20, 6 March 2026 (UTC)

@Glrx Only if the photograph/image is linked to the separate page (similar to Wikipedia), having a link to a credit page buried isn’t acceptable attribution.

I won’t yet support adding Tupungato as a copyleft troll, the fact you haven’t said what this business/site is but being rather vague but regardless if a small, medium or large business, should know how to follow rules and failure to follow them has consequences.

This isn’t someone’s personal site or blog. Bidgee (talk) 22:05, 6 March 2026 (UTC)

@Bidgee: For clarity, this micro-business is three people, with no in-house legal expertise or copyright staff. Do you consider it acceptable for uploaders to use this free CC platform for distribution, then hire Pixsy to extract £900 from operations of this scale for promptly corrected good-faith mistakes?

Standard Pixsy £900/IPEC demand despite fix. Exactly "refusing to allow violators to fix errors rather than pay a fee".^[2]

Tupungato demanded payment from micro-business despite prompt correction. Textbook trolling.

Pomme7755 (talk) 22:39, 6 March 2026 (UTC)

--Pomme7755 (talk) 22:39, 6 March 2026 (UTC)

Then the business should be hiring (when required) a lawyer that deals with copyright and trademark so it doesn't make mistakes that could cost them more than £900!

@Tupungato, can we get your side of the story. Bidgee (talk) 22:47, 6 March 2026 (UTC)

@Bidgee: Expecting 3-person micro-businesses to retain copyright lawyers for Pixsy automation notices = exactly why Commons guideline calls this "trolling".

Tupungato's side: Pixsy auto-demand → business removes image → Pixsy demands £900 anyway → no grace despite fix. Guideline definition verbatim.^[3]

Pixsy defenders here seem invested. Does Commons policy protect such demands against 3-person businesses?

This isn't "business should lawyer up"—it's Commons tolerating Pixsy business model. Pomme7755 (talk) 22:55, 6 March 2026 (UTC)

I don’t have the full story and you’re very vague on it. From now, please do not ping me. Bidgee (talk) 23:18, 6 March 2026 (UTC)

Looking forward to hearing Tupungato's side of the story. If they don't have any clarification to offer, I'm happy to support taking action regarding their images and/or account privileges. Nosferattus (talk) 23:23, 6 March 2026 (UTC)

@Nosferattus: Thanks—timeline matches textbook trolling exactly. First test case for guideline enforcement.

@Tupungato: Your response to not offering attribution fix before Pixsy £900 demand after micro-business removed your single image?

Action needed:

• Troll list addition
• Category:Images by Tupungato deprecation
• Watermark/block review

Pomme7755 (talk) 23:48, 6 March 2026 (UTC)

Hi! Thank you for the ping. I did not realize there was discussion of "copyleft trolling". To begin with, yes, I have account with Pixsy. They have my photos from my self-hosted websites, but also those CC-BY from Wikimedia Commons. Pixsy pursued some of copyright infringement cases for financial resolution, including some of CC-BY from Wikimedia Commons.

Let's start with naming me a "Copyleft troll". I have engaged in editing multiple Wikipedia projects for over 20 years. I have edited articles in English Wikipedia, created articles in Polish Wikipedia, submitted my photos to Wikimedia Commons, patrolled and categorized files in Wikimedia Commons, edited Wikidata. I do this for betterment of the sum of human knowledge and for my love of all Wikipedia projects.

I specifically modified my upload template to underline the fact that to use images outside Wikipedia projects requires attribution. I enabled multiple options for attribution to either my real name, or my website, for potential reusers to have a choice whatever fits them. In some files I even put "Creative Commons Attribution" in Commons filename, to make it additionally visible. I always approached copyright infringement generously. I never enabled Pixsy to pursue any use that attributed to anything reasonable. If the infringer credited to Wikimedia Commons, or Wikipedia, or "Commons", or whatever that resembled any attribution, I never enabled any infringement case for legal resolution. If the infringer credited on general credits page, didn't include the license, didn't include the CC links, I never enabled any case for legal resolution.

One of my popular Creative Commons Attribution photos from Commons is reused about 500 times outside Wikimedia projects. Maybe 10 of these reuses provide correct attribution, about 30-40 provide any attribution at all, 450 of the websites blatantly infringe the copyright and don't care. I enabled Pixsy to pursue only few most blatant infringements. I did not reach out to the infringing company to correct the attribution infringement. After 20 years of doing this I'm only frustrated by usual lack of response, or some sort of cookie cutter response that I need to use their DMCA internal form with formal declaration of rights and original copy of my image straight-from-camera. Companies create massive blog spam for SEO, with thousands of pages stealing thousands of photos. And yet, supposedly I'm the troll and need to defend myself. By the way, why is the infringing company website not named here, so other users would be able to look at this independently?

Anyway, to sum up, I never intended to be a "copyleft troll". I do everything for the spirit of Wikimedia projects, but I'm bitter about blatant constant repeated copyright infringements. I'll accept guidance about correct conduct. Should I never react about no attribution in CC-BY reuse? Tupungato (talk) 17:35, 8 March 2026 (UTC)

@Tupungato: Thanks for the detailed reply and for your long-term work across the projects. I don't doubt your frustration with blatant, no-attribution reuse by large SEO farms and similar.

Where this case worries me (and I think others here) is not the idea of enforcement as such, but the pattern the guideline explicitly calls out as copyleft trolling: "refusing to allow violators to fix their errors rather than pay a fee". In this instance my friend's micro-business had a direct link to your photo on Commons when that article was put together, which was lost during a site migration when parts of some posts were not copied fully, including yours. They are not aware of this discussion here and they don't know about these rules, but I know Wikimedia a bit more and it was strange to me when I heard that they received a first contact that was already a £900 Pixsy demand with IPEC threats, with no prior attempt to request correction. That feels very different from going after a big company that has ignored multiple polite emails.

A few concrete points I'd like to hear your thoughts on. Do you see any role for a "contact first, enforcement later" expectation for obvious good-faith or technical mistakes, especially from very small organisations, before a case is enabled in Pixsy? Would you be willing, going forward, to commit that for cases like this (small reusers, fixable attribution, non-central use) you first send a short, direct note asking for proper attribution, and only consider financial claims if they ignore that?

For what it's worth, nobody is saying "never react to no attribution" or that you have to tolerate large, bad-faith commercial misuse. The concern is about Commons-hosted, "free CC" images being used as inputs to a process where the first interaction with a good-faith reuser is a high, non-negotiable payment demand, with no option to fix. That's exactly the kind of scenario the copyleft trolling guideline was written to address.

If you're open to adjusting your approach around small reusers and fixable attribution, I suspect that would go a long way toward reassuring people that this is about proportionate enforcement rather than turning Commons into a revenue stream. Pomme7755 (talk) 18:54, 8 March 2026 (UTC)

Honestly, with small reusers I wouldn't even send a message, because in my heart they have my blessing to reuse without attribution. I understand small business struggle and I wouldn't enable legal action against such. I won't name any names, but massive media conglomerates taking over small media are among biggest offenders, and "we took over this local newspaper, migrated all the data to our platform and it's not our fault that your image was rehosted without attribution in 800 websites across our syndicated network" is a common excuse. It's never my intention to make small business owners lose sleep. It's not my intention to become "copyleft troll". Tupungato (talk) 19:09, 8 March 2026 (UTC)

@Tupungato: Thank you for clarifying that. It's very helpful to hear you say that small reusers effectively have your blessing, that you would not enable legal action against them, and that you do not want small business owners losing sleep over this.

Given that, can I ask two very concrete things:

1) In this specific case, would you be willing to tell Pixsy that you do not wish to pursue any claim against this micro-business, and ask them to close the file so they stop sending further demands? At the moment, they are still receiving demands for £900 and ongoing IPEC-style threats even though the image has long been removed.

2) Going forward, would you be comfortable making it an explicit practice that you do not authorise Pixsy (or similar services) to pursue small reusers where there is a clear good-faith or technical attribution mistake, and instead treat those cases as "blessed" or resolved once they fix the issue or remove the image?

That kind of clear commitment, combined with what you've already said here, would go a long way towards reassuring people that Commons-hosted CC images are not being used as a lever to put small, good-faith reusers under disproportionate legal pressure, while still allowing you to enforce against the "massive media conglomerates" you mentioned. Pomme7755 (talk) 19:19, 8 March 2026 (UTC)

From this back-and-forth, it sounds like Tupugnato doesn't typically intend to go after small businesses, so either (a) the business we're talking about is not actually very small, or (b) Tupugnato overlooked that it was a small business. That seems reasonably easy to sort out, no? At some point one of you will need to tell us what the business is if you want us to look into this more. If there's a bunch of evidence that Tupugnato is misrepresenting his approach above, we can talk about it at COM:ANU (as a reminder, this is a page for discussing the guideline, not for enforcing the guideline). If there's evidence the nature of the business is being misrepresented, we can all carry on. Absent either of those, however, I don't know that there's anything else we can do here. — Rhododendrites ^talk |  02:22, 9 March 2026 (UTC)

@Rhododendrites @Enyavar: Thanks both. The small business is genuine 3-person micro-operation with no legal dept. Privacy respected as they don't know this discussion exists (don't want to involve them to avoid unnecessary stress while Pixsy demands continue). Standard VRT practice for unaware reusers. Private verification available via VRT if needed.

Tupungato's CC3.0 + Pixsy combo = deliberate choice enabling no-fix £900 demands (as happened here). His "small reusers have my blessing" + "won't enable legal action" directly contradicts this case still active.

Waiting on his response to:

1) Instruct Pixsy close this file (ongoing demands post-removal)

2) "Contact first" commitment for small/fixable cases

If no action/commitment, this validates guideline enforcement (deprecation/listing). Thoughts? Pomme7755 (talk) 03:03, 9 March 2026 (UTC)

@Rhododendrites @Enyavar

@Tupungato:

URGENT UPDATE: Rather than Pixsy's contacts stopping after your reassurances of "small reusers have my blessing" and "won't enable legal action", I just heard that Pixsy called AND emailed this small business today — despite immediate image removal weeks ago.

Tupungato: Please instruct Pixsy to close this file immediately to stop ongoing contact.

A "contact first" policy for small/fixable cases would prevent future incidents.

Please confirm by COB today. This small reuser continues to be targeted otherwise.

Thoughts? Pomme7755 (talk) 14:59, 9 March 2026 (UTC)

Do we get to learn the name of the small business you represent? Once that happens, we should know all groups that are in play here. Enyavar (talk) 15:13, 9 March 2026 (UTC)

@Enyavar thanks for the question. VRT offer stands per standard practice for unaware reusers.

The business is a genuine 3-person micro-operation receiving Pixsy calls/emails today despite weeks-ago image removal.

@Tupungato: Please instruct Pixsy close this file by COB today to stop ongoing contact. Pomme7755 (talk) 15:40, 9 March 2026 (UTC)

I had to Google what COB is. This is my hobby, I'm usually here only during spare time. Small reusers do have my blessing, and I declare "contact first" commitment when possible (sometimes it's almost impossible to find contact info).

Pixsy terms of use mean that photographer authorizes them as agent dealing with legal stuff. To cancel the case photographer must pay Pixsy's expenses to this point, I believe the minimum is set at $200. I'm not willing to add spending $200 in a situation when my copyright was infringed.

I don't know what VRT can do here (I'm not 100% sure what their function is), but I can provide any details privately. Still I consider it uncivil naming me "copyleft troll" from the very beginning, and keeping the other party's details completely secret.

--Tupungato (talk) 18:00, 9 March 2026 (UTC)

I suppose the question is, if it's a tiny business, what went wrong to get to the point of opting to have Pixsy pursue it? Another question, if it was a mistake, is whether they could send you the $ to cancel it rather than the full amount (just thinking outside the box). But I agree the information asymmetry is not good, and I'm having a hard time thinking of a reason why it wouldn't be appropriate to share the information publicly if you're disputing Pomme7755's claims. As they said, they aren't actually a party to the dispute but rather a friend of the party, so it would not be outing. That said, it we want to respect their privacy, I'm also happy to take a look if you want to email information to me, and then give me evaluation on-wiki. That ~$200 fee to cancel is an interesting bit of information regardless, and probably makes sense to document somewhere, as it makes it even more complicated to resolve these disputes. :/ — Rhododendrites ^talk |  18:08, 9 March 2026 (UTC)

I can name the company here as well if the community considers this a good choice, but Pomme7755 seems to avoid it. I don't want to be the first to name it to avoid public shaming or whatever. It's a British limited liability company, so the financials are publically available.

What went wrong to have Pixsy pursue it? At this point I'm not certain. Every few months I'm getting personal message from Pixsy account manager about a case or cases added to a folder with promising cases for monitored images marked personally by Pixsy staff. I was repeatedly assured by Pixsy that no small companies are pursued (which I requested). Tupungato (talk) 16:15, 10 March 2026 (UTC)

@Tupungato Appreciate the nuance on Pixsy's folder system and your "no small companies" requests.

Yet this micro-business—despite immediate removal and prior attribution (lost in migration)—remains pursued months later, wasting time small businesses don't have. Having looked at their public Companies House financials, it couldn't be further from a conglomerate.

Pixsy complaints routinely cite small businesses hit despite such assurances. Photographer discretion is great, but no small business is immune from their automation. No need to name them publicly—VRT remains the privacy-respecting route. Thanks for documenting the CC-BY+Pixsy mechanics. Pomme7755 (talk) 22:26, 10 March 2026 (UTC)

if Pixsy assured you they skip small businesses, perhaps raise this case with them? They might honour their promise by canceling without charges. Pomme7755 (talk) 22:30, 10 March 2026 (UTC)

but the pattern the guideline explicitly calls out as copyleft trolling: "refusing to allow violators to fix their errors rather than pay a fee" - The guideline names this as one of the common themes, and that At what point enforcement begins to look like copyleft trolling will require judgment on a case-by-case basis, but will typically involve some combination (of those common themes). I'm not judging one way or another; just saying that one trait doesn't automatically mean "copyleft troll". — Rhododendrites ^talk |  02:16, 9 March 2026 (UTC)

@Bidgee: Not vague: 3-person business gets Pixsy notice → no ability to fix attribution → removes images → Pixsy demands £900 → no grace. Verbatim trolling definition.^[4]

Others reading: this pattern repeats. Commons needs troll list + deprecation now. Pomme7755 (talk) 23:24, 6 March 2026 (UTC)

Please STOP pinging me! Bidgee (talk) 23:34, 6 March 2026 (UTC)

I cannot stand these dangling ref-links at the far end, so I move them here. --Enyavar (talk) 21:40, 8 March 2026 (UTC)

1. ↑ Commons:Copyleft_trolling. "refusing to allow violators to fix their errors rather than pay a fee"
2. ↑ Commons:Copyleft_trolling. "refusing to allow violators to fix their errors rather than pay a fee"
3. ↑ Commons:Copyleft_trolling. "refusing to allow violators to fix their errors rather than pay a fee"
4. ↑ ...

I am very much not a fan of Commons users having Pixsy send demand letters, and when aimed at anyone other than perhaps a large company or perhaps serial violators, I find it incompatible with the spirit of this project. That being said, this community has had a tough time figuring out what to do outside of extreme cases where intent to profit by uploading here is not clear. We haven't really been willing to do anything about active contributors here who also contract with Pixsy to do the dirty work. There are a range of reasons why, from mere lack of consensus on how exactly to approach such a problem to disagreement on ethical grounds and questions about practicalities of implementation (e.g. we accept imported content where the photographer is not involved and cannot tick a box saying "I agree not to enforce copyright in cases of x y z". On the specifics on the violation, if you use one of Tupungato's photos and just say "by Wikimedia Commons" or something (if that's what happened), that's not a minor mistake but a total absence of attribution. If proper attribution was a link away, we should indeed ensure Tupungato is aware of their mistake and ensure they correct it. Pixsy may take care of the legal stuff, but they only do so after the copyright owner looks at a violation and confirms it's a violation to be pursued. — Rhododendrites ^talk |  04:43, 7 March 2026 (UTC)

== Civility and good faith before Pixsy escalation ==

I would like to comment on this case from the perspective of Commons' stated values around civility, collaboration and good-faith dispute resolution (see Commons:Civility).

The attribution link originally existed but appears to have been lost during a migration

The reuser in this case is a friend of mine and is not aware that I am raising this here. When I asked them about the article, they confirmed that the attribution link was originally present within the article and linked directly to the image on Commons.

Following a site migration in late 2025, parts of several articles were unintentionally lost. Earlier versions of the page visible via the Internet Archive clearly show the attribution link present.

A simple civil interaction could likely have resolved this quickly

Commons' civility expectations encourage raising issues politely with the reuser where possible before escalating further. A short message such as:

"The attribution link appears to have disappeared during migration – could you restore it?"

would very likely have resolved the issue within minutes.

Instead, the issue appears to have been escalated immediately through an external enforcement service. When a straightforward good-faith correction is possible, that level of escalation is difficult to reconcile with Commons' collaborative norms.

If escalation without prior contact is considered appropriate here, it would be useful to understand why the normal good-faith contact step was skipped.

Commons has already faced problems with copyright bounty hunting

Commons has previously had to address the problem of contributors using external enforcement services to pursue settlements rather than encouraging attribution fixes. The Commons:Copyleft_trolling guideline exists precisely because this type of behaviour has appeared before – where enforcement incentives start to resemble copyright bounty hunting rather than protecting the integrity of attribution.

Immediate escalation via Pixsy, without first attempting reuser contact, risks fitting the same pattern that the guideline was written to discourage.

Behaviour has consequences under existing Commons guidance

Commons has already recognised copyleft trolling as a problem. When users act in a way that resembles trolling – prioritising enforcement or settlement over straightforward attribution correction – they should expect the consequences associated with that classification. If someone behaves like a troll, they should be treated as one under the existing community rules and guidance.

Protecting Commons' collaborative mission

Commons works best when reusers experience a cooperative community that encourages attribution and correction in good faith. When the first interaction a reuser receives is an enforcement notice from a commercial service rather than a simple request to fix attribution, that undermines trust in the ecosystem.

Adversarial enforcement should be the last resort, not the first step.

Thoughts? Pomme7755 (talk) 10:27, 7 March 2026 (UTC)

There is a difference between the spirit of what you are saying, which is broadly easy to agree with, and the practicalities of enforcement on Commons for things that do not happen on Commons and which Commons has no legal authority to prevent. If the attribution existed for a while but then did not exist, that's a frustrating mistake, but not possible for anyone other than the site owner to appreciate. Even Pixsy is not, I don't think, constantly reloading sites that include attribution to wait for the moment when they no longer do -- they just look for uses without attribution. I don't have a good answer, but I would be curious to see if there's a pattern with the same user and would like to hear from them if possible. As an aside, please try not to use chatbots to produce talk page messages here. I understand it's a complicated place with various policies, etc., but (a) they often misunderstand the policies (like the civility policy above), and (b) you'll lose a bunch of people who might've responded but don't want to talk to chatgpt (or whichever). FWIW. — Rhododendrites ^talk |  12:46, 7 March 2026 (UTC)

I am sympathetic, but Commons is not a place to resolve an external dispute. Commons does not have access to the evidence, and Commons has no enforcement powers. I wish Commons would add a right to cure to its terms of service, but I do not want such terms to be abused. Nor do I want contributors here needing to patrol the rights of their contributions. I also want all CC files to have the appropriate CC metadata. I do not like Pixsy going after the little guy, but there is some public good in what Pixsy does. Glrx (talk) 17:10, 7 March 2026 (UTC)

Wow. Good to know that there is still some desire to do this.

Some agreements are in order: Rhododendrites is right, I don't want to read chatpot-structured texts. Some of my longer answers in discussions take hours to formulate, and I dislike the superficiality with which chatpots treat the rather complicated matter at hand (In this case, my timer for the two rather succinct posts in this thread today from my keyboard: 95 minutes). But I also agree to the general outrage presented by Pomme7755.

Some disagreements are also in order. Rhododendrites, Commons advertises publicly as a great archive full of material that everyone can re-use. Commons does not advertise as being a legal trap, and yet that is what the trolls are using it for. You wrote above (yesterday 7th, 03:43) about a "total lack of attribution"... and yes, that is why Commons's re-users should be allowed to fix their attribution when they thought it was a free image or when they were mistaken that a comparatively flimsy attribution was sufficient. More in my response to Tupungato below. --Enyavar (talk) 21:34, 8 March 2026 (UTC)

should be allowed to fix their attribution I'm ambivalent on this. I agree with it in spirit, and wonder how it would work in practice. According to this, Walmart could use your photo in an ad, crediting someone else entirely, as long as they fix it if you notice it and ask for a change. Or a newspaper runs a big front-page story illustrated by your photo, and by the time you notice 95% of the readership has already seen it. That doesn't seem great, either. And if we agree that's not ideal, we start to enter the thorny gray area. We don't do a good enough job at telling people what their obligations are when they use things here, and that's something we should be changing, because we fundamentally can't control how people enforce copyright -- even in those extreme cases the best we can do is reduce the odds the most unethical enforcers will do so again in the future. — Rhododendrites ^talk |  02:10, 9 March 2026 (UTC)

Re Commons advertises publicly as a great archive full of material that everyone can re-use - unfortunately this seems to give the mistaken impression of a copyright-free zone, hosting material can be re-used with no conditions, resulting in incorrect re-use to an astonishing degree (see the numbers in Tupungato's post above). It seems like many re-users don't understand that incorrect re-use of Commons media is copyright infringement (just like incorrectly re-using a file from Getty). Is the Commons interface and approach to explaining licensing/license tags not effective? Or is this just the nature of hosting media on the internet? -Consigned (talk) 01:52, 12 March 2026 (UTC)

It's just my theory, but I think it is assumed that Wikipedia is free, therefore everything in Wikipedia is free, therefore Googling for "toilet photo wikipedia" gives you results with toilet photos that are free to use, no strings attached. Some re-users might save the images directly from search results on this assumption. I started putting the phrase "Creative Commons Attribution" in the file name in some of my uploads to avoid this.Tupungato (talk) 17:07, 12 March 2026 (UTC)

I responded above.

The idea that "A simple civil interaction could likely have resolved this quickly" is an empty promise. Typical situation is that infringer doesn't respond at all, responds with hostility, responds with THEIR demands (I need to supply my original straight-from-camera file, I need to supply the scan of my ID/passport, I need to register my work with U.S. Copyright Office).

Tupungato (talk) 18:00, 8 March 2026 (UTC)

@Tupungato, yes, misattributions are a problem, I fully understand and appreciate what you wrote above (today 8th, 17:35). But there are two things: First, you deliberately and to this day upload images under CC-3.0-Unported (!) when the standard choices offered for Commons uploaders are CC-0, CC-BY-4.0 and CC-BY-SA-4.0. That means, you are deliberately choosing this license to predate on re-users. I can imagine no other reason to not switch over to 4.0, when that license is a standard here - and for over ten years now!

The second issue I have is that Pixsy apparently does not offer contracts where they first allow your re-users a short time window to fix the attribution (as would be required by 4.0). If Pixsy would allow "protecting" 4.0-licensed content and thus allow re-users to fix the issue in that time window, then I'd have no problem here. Have at them: go after the re-users who either don't care to fix the attribution, or who apparently have the money to spare.

But: by your actions YOU are are in a position to indiscriminately and without warning confront unsuspecting re-users with 900€ bills. By choosing the license and the enforcer, it is solely YOU who punishes re-users you do not think are legitimate, and YOU who extracts money from them. Are they all sprawling media conglomerates who predate on your content? Right now, it is your say-so against that of your accuser - and yes, they should state their name and website (or allow someone from VRT to verify claims). But as it stands, the "typical situation" you complain about is rather reversed: As the Copyleft Troll, you don't need to respond to any desperate pleas: They pay up, or they pay up a negotiated reduced sum, or Pixsy sues them on your behalf. That may satisfy you and the owners of Pixsy, but the re-users are completely screwed over and the process is totally intransparent. --Enyavar (talk) 21:36, 8 March 2026 (UTC)

as would be required by 4.0 As has come up here a few times before, that seems to be a common misunderstanding of 4.0. You do not erase the violation when you make a change, you cure the license (as in, from that point forward, you can continue to use it legally, but you're still on the hook for the original violation). There is a relevant reason why one wouldn't use 4.0 for cases like this, though: Pixsy won't work with you if you use 4.0, for a variety of reasons including that cure clause. — Rhododendrites ^talk |  02:10, 9 March 2026 (UTC)

@Enyavar The only reason I'm using CC-BY-3.0 is the fact that I contributed images since 2006. There were almost no tools in Commons back then, and I'm still using the basic upload tool, because this is the workflow I learned. The only part of my "deliberately choosing" this license is reusing the basic upload template since around 2008. I didn't realize the differences between 3.0 and 4.0 and from now onwards I can upload 4.0 if it's better

I agree it's my word against someone else's, and I agree that it might be not transparent, about me deciding who I let Pixsy pursue. I'm open to creating some process to do this the right way. Should I not pursue infringement at all? At what point it's ok? You can easily reverse search my images and see how frequently the results have attribution. I'm not sure about the exact number of reuse of my photos with no attribution, but I estimate them to be between 5,000 and 20,000. Sometimes there isn't even easy way to find out who am I supposed to contact to reasonably notify them about infringement. The demand amount was decided by Pixsy (I think based on how many images are on the site, how prominent the image is etc.), but 900€/900GBP is not typical, at least not from the cases I had. 150-400€ is more typical. Tupungato (talk) 17:18, 9 March 2026 (UTC)

@Tupungato: Thanks for the licensing history.

Two questions remain unanswered:

1) Will you instruct Pixsy to close this specific file and stop contacting this small business? (ongoing despite weeks-ago removal)

2) What does your "contact first" process look like for small cases?

@Rhododendrites @Enyavar: Ongoing non-response to these concrete asks, transparency concerns remain. As of today, Pixsy is still contacting this small business despite weeks‑ago removal. I’d appreciate your views on what minimum commitments we should expect here (closing this file + “contact first” for small cases) and at what point deprecation/listing becomes appropriate if that doesn’t happen. Pomme7755 (talk) 17:48, 9 March 2026 (UTC)

No need to keep pinging us. As I said elsewhere, there's nothing we can do to stop Pixsy; even the most extreme action we can take is to ban someone and delete their files, but that has no effect on current enforcement. Tupugnato's responses are also more encouraging than the responses we've heard from others in the past, and I don't think we're anywhere remotely close to taking such extreme actions. I'm also not prepared to do anything else while no evidence has been presented by anyone. And again, this is not the page where we decide whether to take action against someone -- that's COM:ANU, and given Tupugnato's responses here I wouldn't imagine there would be consensus for anything. I hope, if the business truly is very small, that something else can be worked out, but since I have no idea what I'm commenting on, I think that's the last comment from me for now. — Rhododendrites ^talk |  17:55, 9 March 2026 (UTC)

Understood, thanks for clarifying your view of the process.

As Enyavar noted: "Commons advertises publicly as a great archive full of material that everyone can re-use. Commons does not advertise as being a legal trap, and yet that is what the trolls are using it for."

This thread documents those concerns + Tupungato's response, including his "small reusers have my blessing" pledge and future CC-BY-4.0 commitment, for future reusers facing Pixsy enforcement.

We'll see if those words carry substance. Pomme7755 (talk) 18:20, 9 March 2026 (UTC)

Please assume good faith, such comment above is not in the spirit of the project. Bidgee (talk) 18:29, 9 March 2026 (UTC)

@Enyavar: Thanks. As mentioned earlier, the reuser is a micro-business with no legal department. They had attribution initially (link in article), lost during site migration, removed image immediately weeks ago upon notification, yet still receive Pixsy demands to this day. For privacy reasons they are not aware of this thread, which is why I’ve consistently pointed to VRT as the standard route for any private verification.

My main concern is not “outrage” in the abstract but this concrete pattern: small good‑faith reusers of a single CC‑BY image being drawn into automated Pixsy processes with high opening demands and a non‑trivial cancellation fee. This wastes precious time small businesses need for actual struggles, far from Commons' "great archive everyone can reuse" promise.

It’s useful to have on record that (a) Pixsy cases rarely go to court and require the photographer’s additional consent, and (b) photographers can choose to pay Pixsy to disengage "unbearable burden" cases. That confirms how much discretion sits with the photographer in these CC‑BY + Pixsy arrangements. Pomme7755 (talk) 16:34, 10 March 2026 (UTC)

Thanks Tupungato, for explaining your choice of licensing, I think I can understand your position a bit better now. Your engagement is much more helpful than what I remember from Diliff, and please forgive my assumptions where they did not take info account that you simply kept reusing the same template. Myself, I was also unaware of the crucial legal differences between the various licenses, until the Diliff case a few years back, too. Although... at some point you chose to switch from 2.5 to 3.0, at least.

Towards Pomme7755, I've not read up on the long discussion since my previous response, but I think you ought to disclose which image is the one in question as well as the website of the people you represent here - if not to the community that participates on this page, then at least to another volunteer at VRT, who might be able to check and confirm your claims. While I understand the general outrage against learning that people are sent bills for naively re-using Commons content (with no/incomplete/misinterpreted attribution); my outrage goes down a notch when the re-users are less innocent individuals but rather profiteers themselves. By now, I learned more about Tupungato's position, while I cannot say the same for the other side. --Enyavar (talk) 15:21, 10 March 2026 (UTC)

Information about Pixsy policies and procedures

@Tupungato - Thanks for your helpful responses. I'm glad to hear you are not trying to target small good-faith reusers, and it sounds like you are doing your best to live up to the spirit of the project and be fair in your enforcement of your copyrights. The one thing I'm still concerned about is your inability to disengage Pixsy in cases where mistakes are made. If Pixsy is not disengaged from the case, what actually happens? My assumption is that Pixsy just makes endless legal demands until the end of time but doesn't actually take people to court. Do you know if that's true or not? And if Pixsy does actually take people to court, do they have to get your authorization before they take that step? Any insights you could offer about what Pixsy's policies and procedures are in these cases would be very helpful for when we have to deal with actual copyleft trolling. Nosferattus (talk) 22:27, 9 March 2026 (UTC)

I mean, I'd be willing to spend the money requested by Pixsy to disengage the case in accordance with terms of use, if a case turned out to be an unbearable financial burden for the offender.

According to Pixsy guides for photographers "Most cases in our Legal Resolution Program are completed within 2-8 months from the time that a demand letter is sent." In my limited experience no case went to court, and no case was recommended by Pixsy for lawsuit. To proceed to lawsuit, Pixsy is supposed to contact the photographer and receive additional consent. There are actual law firms named in communitation from Pixsy (different for different jurisdictions), so probably lawsuit isn't impossible, but I think unlikely in a Creative Commons case (and would have to be first recommended by Pixsy, and afterwards additionally enabled by photographer).

Tupungato (talk) 14:01, 10 March 2026 (UTC)