I appreciate your intentions, but your examples are just not up to the standard needed to treat AI generated nudes any differently than a nude magazine collage with kids’ crushes faces in it.

As uncanny as the nudes might be, they are NOT accurate. People know this and they are going to learn one way or another to adjust their definition of “real”. No character details like moles or their actual skin tone, or anything like this will be accurately portrayed. They have no reason to think “someone has seen their naked body”. Yeah, if someone tells them to worry about it, they will, as any young person will, but why? The bigger the deal we make of it, the worse it is, and the litmus test is, is it bad if we decide to ignore it and teach kids that ai generated nudes have nothing to do with them and that they can safely ignore them, then they do basically zero harm.

How is your test example related to this at all? In the one case, my face and clothed picture is acquired likely with my implied permission from social media and modifications that i did not authorize are added to it and it is then distributed, making me look naked and having no bearing on my person or character (since the ai doesn’t actually know what i look like naked) so no conclusion anyone would draw from it constitutes a disclosure of information about me. The test example constitutes a clear disclosure with provenance to establish the validity of the information, quire a different scenario. It is true that AI chat bots can be jail-broken to release my previous questions which might reveal things about my character that i do not wish to disclose, but that is a different issue and unrelated to these nude generators.

I’m not saying handing these nudes to a kid or blackmailing them is not criminal or harassment, just that the technology and medium should have almost no bearing on how we treat this.