Isn't the better approach to make deepfake technology so ubiquitous that everyone can plausibly deny that it's really them?
This seems inevitable given the march of technology. Doesn't this also give plausible deniability to real child porn? This makes child porn laws unenforceable in many cases. Unless fake child porn is also illegal. What trade-off do we want there as a society?
Sadly, doubtful. It requires a lot of knowledge and expensive hardware, compared to literally only buying some $20 camera to film.
The people who will win the most from such normalization is people who like CSAM, and people who will lose the most are the victims.
We have libel laws, even though everyone knows the written word can be lies.
This scenario is very much illegal in the US. The bar to clear is photorealistic; if you Photoshop a kid's face onto a nude body, you have in fact produced child pornography. Possessing or distributing that image are also crimes.
*I don't buy the tortured logic in that article that "such production would create opportunities for laundering real wildlife products as cultured products" - if the fake horn is cheaper to make, that's like laundering blood diamonds as coal, at coal's price point.
"Another concern is that the fake horns could drive up the demand for real horns. The status that came with owning a real horn would increase as fake horns become commonly available to everyone" likewise doesn't make sense. How could one tell something is indeed a real horn, and not a fake? And why do fakes make real horns more desirable than they were before? Does Rolex encourage counterfeits, to raise their own prices? It also talks only of "ownership", when horns are used for medicine, not just status. I get the impression they worked backwards from their conclusion to find this logic.
Eventually poaching for real horns becomes too expensive and not really worth the trouble.
The easy availability of generated porn does increase the chances of normalising abhorrent behaviour. AI is also restricted to pictures and writing at the moment, you can't generate videos.
If the demand of CSAM does actually go down, I'm all for AI generated CSAM. However, we have little indication that it does, other than "it makes sense that it would".
Probably because it's one of those things that nobody wants to find out. Especially in countries that have "for-profit" prison systems.
The UK is probably a great example of this. They - the government - fired their drugs advisor because he returned an answer they didn't want to hear.