notjulianjaynes 8 hours ago | next |

It doesnt appear as if the cops generated any sexually explicit AI images of an underage girl. Is this accurate? Is there any legal precedent about AI CSAM yet? While that sort of content is of no interest to me I've been wondering about the ethics of it. How is it different than reading Marquis de Sade or William S. Burroughs (both have written very explicit scenes of violent and sexual abuse of children). Try reading the most fucked up part of Naked Lunch in a Walmart Subway sometime. It's a hoot.

dragonwriter 14 minutes ago | root | parent | next |

> Is there any legal precedent about AI CSAM yet?

Is there specific case law concerning AI depictions? Maybe not, but it is already a federal crime to knowingly produce, distribute, receive, or possesses with intent to distribute "a visual depiction of any kind, including a drawing, cartoon, sculpture, or painting" that (in simple terms) are both obscene and depict or appear to depict minors in sexually explicit scenarios, and there is very little reason to think that courts would carve out AI from this general prohibition.

18 USC Sec. 1446A: https://www.law.cornell.edu/uscode/text/18/1466A

herbst 39 minutes ago | root | parent | prev | next |

What if you instruct the AI that the person is 18 but looks like a child? Is it a child then or is it 18? And who is to judge?

This is going to get interesting in future for sure.

throwaway48540 3 hours ago | root | parent | prev |

Some people argue that you need actual CSAM to produce AI CSAM. I don't think that's true, but in that case, it'd be different because it's a result of abuse, while books are not (directly).