Wikipedia talk:Verifying unusual image licenses

A bit of background

I drafted this proposal after considering the Publicgirluk photo debate, a number of similar situations, and the opinions of various editors. In particular, it became clear that a number of editors (Jimbo included) feel that assuming good faith is insufficient when the images involved are of such an unusual or controversial nature that they might cause harm to others. And that some process was needed to provide Wikipedia with an extra measure for assurance in such cases that the claims being made were genuine.

I believe that this proposal accomplishes that. It balances a need for proof against a desire not to be unduly burdensome on contributors. The goal here is not to provide proof beyond all possible doubt, but rather to provide enough reassurance that a reasonable person could conclude that the actions and claims are most likely to be genuine.

In writing this, I have in part co-opted the existing permissions@ process run by OTRS, as I feel this is the best way to provide the process with both an archival record and reasonable assurances of confidentiality. Dragons flight 19:24, 20 September 2006 (UTC)

The best yet

This is by far the best policy proposal I have yet seen to deal with the issues it deals with. I would vote yes for it as is, but perhaps others have improvements. WAS 4.250 12:42, 20 September 2006 (UTC)

It seems reasonable to me but I still have some questions about workflow. Who will have access to review the replies sent to permissions (at) wikipedia.org? Will that be restricted to Jimbo and the other foundation staff or will that email box be open to general review and confirmation? In one of the paragraphs, the proposal says that "Someone from Wikipedia will attempt to contact the phone number/email address..." Again, will that be foundation staff or is that an activity that volunteer-editors will participate in? If volunteers, how? And how will be ensure that only one person tries to contact them, not a dozen volunteers all trying to help simultaneously and accidentally stepping on each other? Rossami (talk) 13:31, 20 September 2006 (UTC)

Permissions is controlled by m:OTRS staff, a few dozen trusted volunteer Wikipedians from across various projects. The OTRS system, as I understand it, allows specific members of that group to take control of a specific request so that there wouldn't be dozens of people all trying to repeat the same action. Dragons flight 14:25, 20 September 2006 (UTC)

Thanks for the answer. Is that staff willing to take on this additional responsibility? And is the existing OTRS team big enough to handle the expected volume? (I assume that this project will be promoted up from a Wikipedia proposal to a foundation-level process which would mean including all the projects in scope.) If they are willing and able to do it, I think this is a great solution. Rossami (talk) 15:10, 20 September 2006 (UTC)

Look reasonable. HighInBC 16:26, 20 September 2006 (UTC)

My only concerns are that a) such verifications be done in as non-biting a way as possible since image issues already seem to be a way that a large number of new users get bitten and b) That we make very clear to the users in question that the requested personal info is kept confidentially. Apart from these issues which are essentially implementation issues, this looks good. JoshuaZ 18:07, 20 September 2006 (UTC)

Per your first concern, yes I feel that this policy should reflect existing policies of civility and not biting the newbies. While these rules are already in place an explicit mention in this policy does seem warrented. Per you second concern, I also agree that anomnimity to the general public should be preserved. The existing permissions@ system seems to handle this very well by using a few trusted volunteers. Good points, they help the policy without disabling it or making it overly complicated. HighInBC 18:11, 20 September 2006 (UTC)

One more implementation question:
Having been verified, where do you recommend that the OTRS staff member record the verification? The existing proposal says that "Any editor may request verification" either by tagging the page or posting a request to the user's Talk page. If we're going to promise to honor anonymity, how will subsequent editors know that the release for the picture has already been confirmed? Rossami (talk) 19:14, 20 September 2006 (UTC)

We already have {{ConfirmationImageOTRS}}, and I would expect something similar to be added to the image description page once verification had been completed. Dragons flight 19:27, 20 September 2006 (UTC)

Very good idea. Go to it. JesseW, the juggling janitor 02:46, 3 October 2006 (UTC)

Verification

This is a good draft, but this suggestion needs to be tightened:

in a context that clearly references Wikipedia (e.g. with the person holding an "I support Wikipedia" sign or sitting next to a computer showing Wikipedia).

Someone could potentially be tricked into doing this, or have done it, but not for the purpose of granting photo permission. The sign has to specify personal permission and/or GFDL permission, as well as the content (e.g. nude photos of myself). Tyrenius 18:37, 20 September 2006 (UTC)

I agree, the persons feeling towards wikipedia is irrelevent, as an image needs to be available to not just wikipedia. The statement I am User:HighInBC on wikipedia would authenticate any claim or upload made by that user. If User:HighInBC declared an image of himself CC and says he does not mind it being used, then all that is needed is to determine that User:HighInBC is in fact the person in the photo. HighInBC 18:48, 20 September 2006 (UTC)

Agreed with both, and think HighInBC's suggestion is better, since that would cover future interactions with the user, uploading more photos, etc. AnonEMouse ^(squeak) 19:07, 20 September 2006 (UTC)

Agreed. It would mean that the model would have to create an account if they didn't have one, to verify permission, (if the photographer was the uploader) but that's easy enough Tyrenius 19:47, 20 September 2006 (UTC)

It did not occur to me that the model might not have an account. In that case the model could create and account as you say, or use permissions@, which is more conventional in the biz.

Of course, it still needs to be addressed what to do if the model has surrendered rights to the photograph to someone, and that someone posts it legitimatly(proper license, on topic, imformative ect...), and the model wants it removed(with no legal justification or rights to the image). Of course, it may be better to keep that out of this policy and address it elsewhere as it seems to be another topic alltogether. HighInBC 20:05, 20 September 2006 (UTC)

If a model has had explicit photos taken and has signed over the rights, then the model has no rights! Tyrenius 00:49, 21 September 2006 (UTC)

I agree. HighInBC 01:38, 21 September 2006 (UTC)

or sitting next to a computer showing the image on Wikipedia: this would be much easier to fake in Photoshop, so I think should be scrapped. Tyrenius 00:51, 21 September 2006 (UTC)

The goal is not to provide absolute proof, since most any system we could imagine could be defeated by a clever con artist. The goal is to provide a reasonable margin of reassurance about the legitimacy of the images without imposing too large a burden on the uploader. I think a person holding sign or sitting in front of a computer screen are both hard enough to fake that a reasonable person would believe that things are likely to be okay. Keep in mind that these are steps taken merely because an image is unusual and without any other evidence of a dispute. If there is an actual complaint, we would be expected to go beyond what appears on this page. Dragons flight 01:30, 21 September 2006 (UTC)

An image can be faked in photoshop, but you can tell. The best fake digital image can be detected by a studious observer. There are many people on wikipedia who has image knowledge. Sure, a fake may get by, that is true with any verification system, but one day someone will see it and say That is fake!. HighInBC 01:36, 21 September 2006 (UTC)