Or more to the point, “this picture contains copyrighted material, which must be censored because…”

etc.

I tried to be as general as possible.

The training data for a self-censorship neural network could be as robust as any given society would like.

An algorithm based on self-censorship of generated output wouldn’t require censorship within the training data for the generative neural network.

I can imagine some other advantages to that approach.

But literally (and I use that word literally) none of the pictures contain copyrighted material.

I don't know how people can make these strong statements about anything in law.

Disney have won cases in court were some artist has drawn their own version of Mickey Mouse, similarly try writing a story about some kids in a wizard school and you need to be extremely careful not to violate (or at least get taken to court) for Harry Potters copyright.

I'm pretty certain image production models have produced some images which would very likely to be judged to violate copyright (a much less strong statement).

You are confusing copyright with trademark. Or, provide a link showing the images case was decided on copyright issues, and I’ll reconsider my position.