Its a very difficult subject, both sides have merit. I can see the "CSAM created without abuse could be used in treatment/management of people with these horrible urges" but I can also see "Allowing people to create CSAM could normalise it and lead to more actual abuse".

Sadly its incredibly difficult for academics to study this subject and see which of those two is more prevalent.

There is also the angle of generated CSAM looking real adding difficulty in prosecuting real CSAM producers.

The fact that you don't need to actually supply any real CSAM to the training material is the reasoning being offered for supporting AI CSAM. It's gross, but it's also hard to argue with.

Yeah, this is basically the crux of the issue. When you get into the weeds and start looking at more than just surface-level “but it needs CSAM to make CSAM” misconception, arguments against it basically boil down to “but it’s icky.” Which… Yeah. It is. But should something being icky automatically make it illegal, even if there are no victims?

I hate to make the comparison (for a variety of reasons) but until fairly recently homosexuality was psychologically classed as a form of destructive/dangerous kink. Largely because straight people had the same “but it’s icky” response whenever it got brought up. And we have tried to move away from that as time has passed, because we have recognized that being gay is not just a kink, it’s not just a choice, and it’s not inherently dangerous or harmful.

To contrast that, pedophilia has remained stigmatized. Because even if it passed the first two “it’s not just a kink/choice” tests, it still failed the “it’s not harmful” test. Consuming CSAM was inherently harmful, and always had a victim. There was no ethical way to view CSAM. But now with AI, it can actually begin passing that third test as well.

I don't know how I feel about it, myself. The idea of "ethically-sourced" CSAM doesn't exactly sit right with me, but if it's possible to make it in a truly victimless manner, then I find it hard to argue outright banning something just because I don't like it.

This is really the biggest hurdle. To be clear, I’m not arguing that being an active pedo should be decriminalized. But it is worth examining whether we’re basing criminality purely off of the instinctual “but it’s icky” response that the public has when it gets discussed. And is that response enough of a justification for making/keeping it illegal? And if your answer to that was “yes”, what if it could help pedos avoid consuming real CSAM, and therefore reduce the number of future victims? If it could legitimately help reduce the number of victims but you still want to criminalize it, then you are not actually focused on reducing harm; You’re focused on feeling righteous instead. The biggest issue right now is that harm reduction is very hard to study, because it is such a taboo topic. Even finding subjects to self-report is difficult or impossible. So we’ll have no idea what kinds of impacts on CSAM consumption (positive or negative) AI will realistically have until after it is widely available.

Plenty of hentai out there covering questionable subjects to train AI on as well.

I've been kind of on the fence about this but then research found that people who physically or verbally express their anger tend to get angrier or delay calming down, contrary to conventional wisdom. I wonder if there could be a similar pattern with this so now I'm hesitating.

When my dad worked for the DoD, he was assigned a laptop for work that had explicit photos of children on it.