doesn’t it follow that AI-generated CSAM can only be generated if the AI has been trained on CSAM?
This article even explicitely says as much.
My question is: why aren’t OpenAI, Google, Microsoft, Anthropic… sued for possession of CSAM? It’s clearly in their training datasets.
For me, this was at no point about the morality of it. I’ve been strictly talking about the definition of terms. While laws often prohibit both CSAM and depictions of it, there’s still a difference between the two. CSAM is effectively synonymous with “evidence of crime” If it’s AI generated, photoshopped, drawn or what ever, then there has not been a crime and thus the content doesn’t count as evidence of it. Abuse material literally means what it says; it’s video/audio/picture content of the event itself. It’s illegal because producing it without harming children is impossible.
EDIT: It’s kind of same as calling AI generated pictures photographs. They’re not photographs. Photographs are taken with a camera. Even if the picture an AI generates is indistinguishable from a photograph it still doesn’t count as one because no cameras were involved.
Right. I get you, and I agree, and I don’t think Buffalox was contradicting you by essentially saying “even if they technically aren’t the same, your government may still count it as the same.”
Yeah, and I think Buffalox agrees aswell. We were simply talking past each other. Even they used the term “depictions of CSAM” which is the same as the “simulated CSAM” term I was using myself.