- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
Hugh Nelson, 27, from Bolton, jailed after transforming normal pictures of children into sexual abuse imagery
A man who used AI to create child abuse images using photographs of real children has been sentenced to 18 years in prison.
In the first prosecution of its kind in the UK, Hugh Nelson, 27, from Bolton, was convicted of 16 child sexual abuse offences in August, after an investigation by Greater Manchester police (GMP).
Nelson had used Daz 3D, a computer programme with an AI function, to transform “normal” images of children into sexual abuse imagery, Greater Manchester police said. In some cases, paedophiles had commissioned the images, supplying photographs of children with whom they had contact in real life.
He was also found guilty of encouraging other offenders to commit rape.
At least with a human being it’s a matter of factuality whether or not they’re over 18. But with AI it’s unverifiable, especially considering some models have already been trained on CSEM.
Once someone has that model locally, do they technically possess CSEM, even unknowingly? Do they only possess it if they try to make the AI make it? Seems like something someone in charge should have thought about in a legally binding way before dumping the internet into an image generator!
In this case he used pictures from actual children and transformed them into CSAM using AI. So there’s no question about the age, and there are real victims, too.
Oh yeah, this dude without a question is guilty and a pedo. I meant more that ‘out of the box’ models may still produce material that looks really CSEM adjacent, and you have no way of telling whether or not it used CSEM to generate the image if the whole dataset is poisoned by actual CSEM being included.
I assume any CSEM ingested into these models is absolutely swamped by the massive amount of adult porn that’s much more easily available. A handful of images aren’t going to drive model output in datasets of the scale of the image generation models. Maybe there are keywords that could drill down to be more associated with the child porn, but a lot of “young” type keywords are already plentifully applied to adults, and I imagine accidental child porn ingests are much less likely to be as conveniently labeled.
So maybe you can figure out how to get it to produce child porn, but it probably won’t just randomly produce it for an innocent porn prompt.
The actual issue is that models who are trained on porn, or even just nudity, and simultaneously trained on perfectly innocent pictures of children will be able to produce at least an approximation of CSAM if you know what you’re doing.
More recent commercial foundation models are absolutely neutered when it comes to nudity and, or at least that’s the hypothesis, that’s why so they’re godawful at anatomy. Which is why upcoming community models are going to go the way of include the porn but not include any pictures of any child in any situation.
Absolutely agree. My comment above was focused on whether some minimal amount of CSEM would itself make similar images happen when just prompting for porn, but there are a few mechanics that likely bias a model to creating young-looking faces in porn and with intentional prompt crafting I have no doubt you can at least get an approximation of it.
I’m glad to hear about the models that are intentionally separating adult content from children. That’s a good idea. There’s not really much reason an adult-focused model needs to be mixed with much other data. There’s already so much porn out there. Maybe if you want to tune something unrelated to the naked parts (like the background) or you want some mundane activity, but naked, but neither of those things need kids in them.
Bound to be tested in court sooner or later. As far as I understand it one is “in possession” if they have access to a set of steps or procedures that would recover an image. So this prevents offenders from hiding behind the fact their images were compressed in a zip file or something. They don’t have a literal offending image, but they possess it in a form that they can transform.
What would need to be tested is that AI generators are coming up with novel images rather than retrieving existing ones. It seems like common sense but the law is quite pedantic. The more significant issue is that generators don’t need to be trained on csem to come up with it. So proving someone had it with the intent of producing it would always be hard. Even generators trained on illegal material I’m not sure it would be straight forward to prove that someone knew what it was capable of.
I’m not the person to clear up this legal grey area. I just think that AI porn often has these incredibly young faces which makes the enjoyers of that porn extra creepy.
Creeps, or young themselves, and/or have shit all for an artistic eye. AI models have given us a whole new uncanny valley and that’s photorealistic textures on comic/anime abstraction.
Oh yeah, I was agreeing with you, sorry if I was unclear. It pisses me off this situation exists to begin with.