Hacker News Reader

I have one serious issue that I haven't seen mentioned before.

Wouldn't this open up a giant legal loophole for people to claim that ther CSAM is "just Ai generated"?

Yes, mandatory watermarking could be considered a solution ...but it still introduces a giant layer of doubt.

The feds would probably be able to use metadata and facial recognition comparison against existing CSAM databases to verify many such cases. And investigators would likely use hints such as the image resolution or comparison to other images on the drive to verify if a collection is a photoshoot or a crafted prompt for example.

Reply on news.ycombinator.com