A sex offender convicted of making more than 1,000 indecent images of children has been banned from using any “AI creating tools” for the next five years in the first known case of its kind.
Anthony Dover, 48, was ordered by a UK court “not to use, visit or access” artificial intelligence generation tools without the prior permission of police as a condition of a sexual harm prevention order imposed in February.
The ban prohibits him from using tools such as text-to-image generators, which can make lifelike pictures based on a written command, and “nudifying” websites used to make explicit “deepfakes”.
Dover, who was given a community order and £200 fine, has also been explicitly ordered not to use Stable Diffusion software, which has reportedly been exploited by paedophiles to create hyper-realistic child sexual abuse material, according to records from a sentencing hearing at Poole magistrates court.
As a UK citizen, I’m ashamed of my government.
I am firmly against child abusers, but AI images don’t harm anyone and are a safe and harmless way for pedophiles to fulfil their urges, which they cannot control.
Where does the training data come from to create indecent images of children?
It doesn’t need csam data for training, it just needs to know what a boob looks like, and what a child looks like. I run some sdxl-based models at home and I’ve observed it can be difficult to avoid more often than you’d think. There are keywords in porn that blend the lines across datasets (“teen”, “petite”, “young”, “small” etc). The word “girl” in particular I’ve found that if you add that to basically any porn prompt gives you a small chance of inadvertently creating the undesirable. You have to be really careful and use words like “woman”, “adult”, etc instead to convince your image model not to make things that look like children. If you’ve ever wondered why internet-based porn generators are on super heavy guardrails, this is why.
I’m not going to say that csam in training sets isn’t a problem. However, even if you remove it, the model remains largely the same, and its capabilities remain functionally identical.
At that point it’s still using photos of children to generate csam even if you could somehow assure the model is 100% free of csam
That would be true, it’d be pretty difficult to build a model without any pictures of children at all, and then try and describe to the model how to alter an adult to make a child. Is anyone asking for that though? To make it illegal to have regular pictures of children in these datasets?
No but it is a reason why generating csam should be illegal. You’re using data trained on pictures of real kids
Thanks for the reply, it’s given me a good idea of what’s most likely happening :)
It’s a shame that the rest of the thread went to shit, but unfortunately it’s an emotional topic, and brings out emotional responses
Always happy to try and productively add to someone’s learning.
It is true, a 10 year old naked woman is just a 30 year old naked woman scaled down by 40%. /s
No buddy, there isn’t some vector of “this is the distance between kid and adult” that a model can apply to generate what a hypothetical child looks like. The base model was almost certainly trained on more than just anatomical drawings from Wikipedia - it ate some csam.
If you’ve seen stuff about “Hitler - Germany + Italy = Mousillini” for models where that’s true (which is not universal) it takes an awful lot of training data to establish and strengthen those vectors. Unless the generated images were comically inaccurate then a lot of training went into this too.
Right, and the google image ai gobbled up a bunch of images of black george washington, right? They must have been in the data set, there’s no way to blend a vector from one value to another, like you said. That would be madness. Nope, must have been copious amounts of asian nazis in the training set, since the model is incapable of blending concepts.
You’re incorrect and you should fucking know better.
I have no idea why my comment above was downvoted to hell but AI can’t “dream up” what a naked young person looks like. An AI can figure that adults wear different clothes and put a black woman in a revolutionary war outfit. These are totally different concepts.
You can downvote me if you like but your AI generated csam is based on real csam so fuck off. I’m disappointed there is such a large proportion of people defending csam here especially since lemmy should be technically oriented - I expect to see more input from fellow AI fluent people.
You’re spreading misinformation and getting called out for it.
Just a note - csam has been found in model training sets: https://cyber.fsi.stanford.edu/news/investigation-finds-ai-image-generation-models-trained-child-abuse
It isn’t misinformation, though, generative AI needs a basis for it’s generation.
The whole point of diffusion models is that you can generate new concepts using training data. Models trained on any nsfw images can combine those concepts with any of its non-nsfw concepts. Of course, that’s not to say there isn’t CSAM in any training data, because there objectively has been in the past, but there doesn’t need to be any to generate it.
Thanks for the reply, that makes a lot of sense :)
Thanks for not being a dick! I aim to inform
Ai is able to fill in the last field in a table like “Old / young” vs “Clothed / naked” when given three of the four fields.
Csam is in the training data. From a few months ago
Please reiterate your statement but instead using the “goose chase meme” format.
It brings me to ask the question if lolicon could be their next target?
The sexualization of children in any form is disgusting and harmful.
Pedophiles need medical and psychological help, not encouragement in fulfilling their urges. Society should reject the sexualization of children. Full stop.
I don’t know what the right answer is, but we provide substitutes for drug addicts to help them overcome their addictions. Methadone and nicotine patches come to mind.
Is it completely inconceivable that a similar tool would help with harmful sexual desires?
I was listening to a podcast on moral philosophy (wouldn’t you wanna be as cool as me??), and one suggestion that’s stuck with me was the morality of, trigger warning,
spoiler
‘life like child sex robots’.
As in, would we as a society want to permit such things, knowing that they could potentially save humans from actual harm if they offer an outlet that scratches an itch? On the other hand, would they bring forth more harmful desires in a greater number of potential perpetrators, leading to even more harm?
Anyway, I’m glad it’s not my job to contemplate such disturbing topics.
Permit is a difficult thing. There is always someone willing to make such things profit. Yay, capitalism!I
I have no idea where they come from but I’ve dove deep enough into the darker web to know people possess such items. I have no doubt that anyone willing to dive in pursuit of such an item would have little problem in finding one.
I recall reading an article about the creators of the real doll and how they received such requests but refused them. They also claimed they had requests for animals which they didn’t think were serious and also refuse.
I’m sure those exist as well and I would rather those exist than for people to harm real animals.
The underlying behavior is the problem though. While substitutions could potentially be made available, this isn’t the same as drug addiction. The reality is that while a pedo could be satiated with a drop in replacement for a time (and possibly indefinitely), there is a very real risk that after a while they’re not satisfied with pretending and could quickly jump to the real thing in a split second. The due course, in my mind, is either modifying the depraved behavior or removing the person from society. While drug addiction can be a vice that doesn’t inflict harm on the rest of society (ie an addict is potentially able to silo their use from the rest of society), pedophilia is always a crime with a victim. The entire purpose of the situation is ensuring that no one becomes a victim of sex crimes, especially minors, and it is too great a risk to allow in any form.
It also runs the risk of giving pedophiles something to form communities around wherein they reinforce the idea amongst one another that society is the one who has a problem not them.
Additionally, in those communities it becomes more difficult to identify pictures and video of abuse to real children by muddying the water. The more sophisticated AI image generation becomes, the harder it will become for law enforcement that deals with this shit.
Let’s compare it with adult pornography. Does the consumption of adult pornography remove the desire to have sex with another adult in the long term? Or does it reinforce the sexually desirable characteristics of adults?
Well considering porn addiction can often lead to lower libido and decreased performance with a partner, sorta yeah.
Current mental help methods for pedophiles include acceptance of their desires as normal, just not something to act on IRL.
It does not prohibit any fictional materials including children, nor can it make someone uninterested in children.
By stripping away safe outlets, we may come at risk of these people increasingly turning to real CSAM, which is way more harmful.
What do you know of the current methods. Where did this information come from? I’d really like to see it. You spoke with such knowledge, you must have the data to make it up, right?
The approach was originally pioneered by the Prevention Project Dunkelfeld, and later adopted for a wider use in Germany, Europe and abroad.
Studies have shown that this approach does work, which led to its widespread adoption and popularization.
You can read details of the treatments coming out of this research here.
Beware of the corporate greed and prepare good old Sci-hub to read sources in full text if you want to.
i read three of the sources you provided (all of them, except the book), and the only thing you’ve said which is true is that the treatment ‘includes acceptance of their desires’ (though you have added the words ‘as normal’)
the other two claims you’ve made, including ‘it does not prohibit any fictional materials including children’ and ‘by stripping away safe outlets we may come at risk of these people increasingly turning to real CSAM’ are your own inventions, and are not stated anywhere in the texts you have linked, in fact, they are directly refuted by both of them, because the actual prevention project recommends a combination of cognitive behavioral therapy and medication
I specifically addressed the “current methods” part of it, as questioned.
The second point was beyond the scope of the sources I provided, except maybe the book, but the project is in line with this as well - it does not focus on the fictional materials and does not explicitly prohibit them. It doesn’t encourage the consumption of such materials, either, so the position can be best described as “neutral”. It does, however, strongly object real CSAM.
The latter was answered in another thread - yes, you are right about this being my speculation, as the scientific community, for all I know, currently doesn’t have data to either prove or disprove this point. But that seems likely to me.
Current mental help methods for pedophiles include acceptance of their desires as normal, just not something to act on IRL.
I am not aware of the research in this area although I have a minor psych background so that’s interesting and makes sense in hindsight. My understanding is that a large part of the compulsion is driven by guilt, shame, feelings of worthlessness, prior victimizations of themselves, etc. Essentially trying to gain a sense of power by taking it from those more vulnerable than them, like an abuser beating their spouse because someone at work put them down. So it makes sense to encourage a sense of power and lessen any sense of guilt and shame.
On a side note, I can’t imagine having their name plastered everywhere does anything but trigger the compulsion to re-offend. Maybe when we advance more as a society, we can separate individuals into categories of has-offended and child-attracted, with the former being on a public danger list and the latter having frequent discreet visits by social workers and mandatory counselors, etc. To lessen the chance of offense and possibly start helping them before they get to the offense stage (those that were ever going to offend.)
I’m pretty sure that this is not true. I’d love to see sources.
There was some research before the ongoing AI-panic, focusing on hentai instead. As it is as “harmless” as the AI-generated content.
And I do recall that at the time there were voices in research making the point that the consumption of material did not have correlation with actually reducing the urges. So this seems highly unlikely.
Upon proper search, I agree I must have been too rushed in decisions as the topic of the influence of computer-generated or drawn CSAM on escalation in offending still seems to be a matter of speculations, with severe lack of sources on both sides (correct me if I’m wrong).
Both sides draw from singular testimonials.
Still, I will remove the notion on science. Thank you for issuing the correction.
P.S. This paper does some job of evaluating both sides, although has its own strong bias not based on presented evidence. Still, it is useful to get some basic overview of the current state of affairs.
This sort of problem-solving acumen is how HIV became so widespread in Africa. Have you considered instead trying competence?
Have you considered not advocating for the sexualization of children?
You’re using sensationalist language to hide the fact that you’re the person in this conversation arguing for more aexual abuse of children, not me. I’m advocating for using drawn artwork to reduce abuse of children. You’re advocating for banning said artwork, which we know leads to more abuse of real children.
Please stop supporting pedophilia.
UK legislators have a long history of taking actions not informed by science or reason but rather the popular, often hysteric, opinion.
This case is yet another attempt at tightening screws where they shouldn’t be.
AI imagery was produced by Stable Diffusion, the model that, for all we know, did not take real CSAM as inputs and caused no harm to actual children. At the same time, such images are important at discouraging the consumption of real CSAM, with very real children being traumatized.
By banning AI imagery production using safe models, legislators leave no legal way for pedophiles to get something by the harmless means, directing many to the harmful ways as equally illegal, while also prosecuting those who did no harm.
I don’t have any of these Tendencies but I like to think that if I did I would chemically remove my sex drive
That’s up to everyone. Besides, most pedophiles do have sexual interest towards adults as well, and current means reduce that drive too.
Chemical castration in this context increases misery and makes building healthy adult relationships harder. Most pedophiles do not opt for that, for all I know.
Current therapeutic methods do include suppressing sex drive in case the client struggles with impulse control. Otherwise, it is not offered, but can be given on request.
counter point:
if you have a folder of AI generated CP and put in a couple of pictures of actual CP it’s going to muddle the case as the offender could claim all of them are simply AI generated. Real harm could go unnoticed if those two were to be treated differently.
Additionally, not every offender will stop at AI generated images, and if their curiosity becomes enough they could go on to want to experience “the real thing”.
You do realize that slippery slope argument is what’s used when it comes to banning anything else, right?
“Can’t legalize marijuana or people will start wanting to do meth” for example.
I don’t believe those two are comparable.
Weed and meth are rather different in how they affect people.
AI images are often used as a way to imitate reality
It doesn’t matter if you believe it, for those who lived through D.A.R.E and the war on drugs, that argument was common and on plenty of people’s lips. It’s a stupid argument but I think that’s the point OP is trying to make
then why is that person repeating a stupid argument at me? those aren’t comparable at all.
A better comparison would be idk, CBD weed with no THC being legal and that being the “gateway” to normal weed. Or buying a knock off product and wanting to try the original. Or looking at AI generated photos people eating spaghetting and wanting to see how it actually looks like
It’s a stupid argument being juxtaposed with your argument…you’re so close, you got this.
I think the solution here is not banning AI materials outright but to make them identifiable - even by means of digital signatures if you want.
For example, Stable Diffusion could insert particular piece of metadata into the picture containing the signature and proving the image is AI-generated, etc.
Without AI materials, said curiosity may lead people straight to the “real thing”, and every darknet or even Telegram dweller will tell you it’s frighteningly easy to find it even if you never intended to. With AI materials, people can have a chance to stop there.
deleted by creator
The fact that digital signatures don’t work like that and sign specific content.
It’s a mathematical procedure that involves the data being signed. If you apply it to any other content - the sign won’t be valid.
meta data is trivially easy to strip off a picture, you don’t even need to bother using tools for it - just take a screenshot and delete the original
Can be baked in pixels, or even better sent to identification for a system similar to what Apple uses to detect CSAM, but as an “alright” ID (but just in police’s hands, not on device or something).
But even then, if every pixel gets marked as ‘created by AI’, it would still be trivial to take real CSAM and run it through an image-to-image generator with denoising turned down to 0.05 and suddenly you have real CSAM that has been marked as ‘legal’ since it is technically AI generated.
Also, keep in mind that there are several open source projects out there where anyone who knows what they are doing could just strip out any protections that might be put in place.
Apple-like ID system solves the latter by technical means.
As per image-to-image, feeding the model with recognised CSAM should be unavailable to begin with.
Yeah but the point is you can’t easily add it to any picture you want (if it’s implemented well), thus providing a way to prove that the pictures were created using AI and no harm has been done to children in their creation. It would be a valid solution to the “easy to hide actual CSAM between AI generated pictures” problem.
you can’t easily add it to any picture you want (if it’s implemented well
Edit for the downvoters: StackExchange - How do I add exif data to an image?
Going to need you to elaborate on this. EXIF data is just bytes in a file, like any of the other bytes in the file. It can be changed and is often changed without the users consent. Are you proposing we create a new type of hardware, something akin to Secure Enclave, and then mass-produce and add it to every consumer CPU to ensure some specific types of exif data isn’t tampered with?
I was thinking of an approach based on cryptographic signatures. If all images that come from a certain AI model are signed with a digital certificate, you can tamper with metadata all you want, you’re not gonna be able to produce the correct signature to add to an image unless you have access to the certificate’s private key. This technology has been around for ages and is used in every web browser and would be pretty simple to implement.
The only weak point with this approach would be that it relies on the private key not being publicly accessible, which makes this a lot harder or maybe even impossible to implement for open source models that anyone can run on their own hardware. But then again, at least for what we’re talking about here, the goal wouldn’t need to be a system covering every model, just one that makes at least a couple models safe to use for this specific purpose.
I guess the more practical question is whether this would be helpful for any other use case. Because if not, I hardly doubt it’s gonna be implemented. Nobody is gonna want the PR nightmare of building a feature with no other purpose than to help pedophiles generate stuff to get off to “safely”, no matter how well intentioned
How could they possibly enforce this ban?
What do you mean? Like how would they catch him?
In the States parole/probation means you lose most of your civil liberties. In other words, if this was the U.S. a PO would check his phone and possibly his computer. Possibly even pull ISP records depending on how bad they want to catch you/how full of shit they think you are.
How will they even know he’s doing it? It doesn’t say they’re monitoring his internet connection. And even if they were monitoring his internet connection, he could go to some public wifi hotspot and sit in a car and do it.
I edited my comment. You’re too quick.
But yeah, he could get around it. But, he’s an addict. He’s going to want that porn other places then his car and make mistakes. If he’s tech savvy, he can probably stay one step ahead of his probation agent (assuming he has one). If he’s not, he’ll slip up because he’s addicted, and that’s how people get caught.
Is it weird that the whole detect/evade game just sounds super fun to me?
Not really. You’re probably a bit of a dopamine/norepinephrine (adrenaline) junky, like most Westerners. It’s bred into us by consumer culture.
It’s weird that it’s not weird though.
Nahh, I’m not so much about the chase as the metagame.
Might want to checkout cyber security and pen testing. It’s not the same thing exactly but it kinda close in some regards.
What do you mean by metagame? Like you find the cat and mouse stuff that is happening fascinating?
Put monitoring software on his devices.
He could just get a burner phone. Realistically, there is no way to police this.
This is pretty similar to restraining orders, make it more difficult and make the consequences more severe.
In the modern world when we have cellphones that can do pretty much anything… it’s fucking hard. There will be a parole officer and monitoring software with periodic physical inspections along with watching his purchases. (That’s, at least, th American approach).
Usually the way it works is that when this dude slips up once he goes to prison for violating his court order.
Or a burner laptop/Chromebook/whatever. Couple that with a VPN, using a neighbor’s wifi, public hotspots, etc, I don’t really see how they can realistically enforce someone motivated to do what they’re gonna do.
Have part of his probation be having his property searched to check for such devices.
There’s a log for everything. There really is. It’s just hard to piece it all together.
Difficult indeed. Maybe banning him from owning a computer or graphics card capable of running local models. Installing spyware on his devices. I’m sure he could get around it, undetected, for a while, if he was determined. But he would be gambling at getting caught.
To be honest I don’t really see the point. We are getting to a point where anyone can generate any image of anything at any time. Let him have his fun. We can’t sexual conversion therapy him and he isn’t hurting anyone. Ones and zeroes, bits and bytes.
For anyone who thinks conversion therapy works; imagine someone trying to convert you from your sexuality. Let’s say I tried to make you gay. Would that ever work for you?
Maybe they think he is capable of self enforcing the ruling?
Or that they want the option of gaoling him if they so much as get a hint he’s using one of the services in any capacity.
Why didn’t he get banned from using the internet?
Is he extorting actual kids or just having a computer generate fap material? The difference decides whether or not I give a damn.
He is fapping to porn that was generated by an AI that trained on csam.
Yes, just like the pictures of astronauts on horses were trained on an extensive collection of space derby pictures.
Not quite. You see, unfortunately, space derbies don’t actually exist. The other, unfortunately, actually does.
Be in denial if you want. That csam is trained on csam.
Any proof for this? Would be an interesting read.
No, it’d be hard to tell since models are usually close lipped. But Twitter has been included in a lot of the image models and traditionally has a very large issue with cp.
Oh… You sounded so confident at first.
See the sibling comment with a link.
Kind of just contradicted yourself there. And have you ever heard the phrase “correlation does not imply causation”?
But how can that be? Surely just the fact that it can create those pictures is incontravertable proof that it was trained on pictures of spacesuited cowboys?
He used Stable Diffusion, which, for all we know, was NOT trained on CSAM.
Csam is in the training data. From a few months ago
Thanks for the correction!
It’s worth noting that this only includes CSAM accidentally scraped along with everything else on the open Web. No specialized CSAM training took place.
In any way, I welcome the efforts at filtering such content before it enters the dataset.
It’s obviously accidental, but that doesn’t change that it happened and is something that will be near impossible to avoid as long as they continue to scrape data in the way they do for their models. They would need a human to filter it out like they already use for most LLMs.
I definitely can’t let you do that, Hal.
How you will enforce this kind of politics? I just buy a VPN, use proxychains or annonsurf, what you gonna do? put a police to live in the same room as I live?
Install spyware on your devices, one would assume.
I think it would be hard to enforce but it’s an interesting precedent, and it means that if they catch them they can whack him
This is the best summary I could come up with:
The Internet Watch Foundation (IWF) said the prosecutions were a “landmark” moment that “should sound the alarm that criminals producing AI-generated child sexual abuse images are like one-man factories, capable of churning out some of the most appalling imagery”.
Susie Hargreaves, the charity’s chief executive, said that while AI-generated sexual abuse imagery currently made up “a relatively low” proportion of reports, they were seeing a “slow but continual increase” in cases, and that some of the material was “highly realistic”.
The Lucy Faithfull Foundation (LFF), which runs the confidential Stop It Now helpline for people worried about their thoughts or behaviour, said it had received multiple calls about AI images and that it was a “concerning trend growing at pace”.
The decision to ban an adult sex offender from using AI generation tools could set a precedent for future monitoring of people convicted of indecent image offences.
Stability AI, the company behind Stable Diffusion, said the concerns about child abuse material related to an earlier version of the software, which was released to the public by one of its partners.
It said that since taking over the exclusive licence in 2022 it had invested in features to prevent misuse including “filters to intercept unsafe prompts and outputs” and that it banned any use of its services for unlawful activity.
The original article contains 974 words, the summary contains 219 words. Saved 78%. I’m a bot and I’m open source!
wank foundation
Wank watch foundation
