Man used Grok to make 7K sex images of stepdaughter, then shot himself
In March, a girl’s stepfather took his own life after cops discovered that he had used Grok to create 7,000 sexually explicit images using one photo taken when his stepdaughter was 11 years old, the amended complaint alleged.
Grok allowed the man to generate extreme images depicting incest and rape without flagging any harmful behavior, the complaint said. Seemingly, xAI’s child safety system only intervened after the man input a prompt for “gang rape.” That request sent a CyberTip to the National Center for Missing and Exploited Children (NCMEC), which alerted law enforcement to the AI CSAM.
Yet the harm was not stopped then, either. Despite mandatory reporting requirements to share information like a user’s IP address when CSAM is flagged, xAI repeatedly refused to help cops or NCMEC identify the user, the complaint alleged. For weeks, xAI allegedly “obstructed this investigation at every turn” and made it harder for “law enforcement efforts to locate, identify, and apprehend the perpetrator.”
Eventually, the stepfather was arrested after cops obtained a warrant to seize his devices. That’s when “a forensic review revealed approximately 7,000 AI-generated images and videos” depicting his stepdaughter, which were allegedly produced using Grok. Without Grok providing users with easy access to “undressing” capabilities, his family doubts he ever would have generated the harmful images, which he allegedly trafficked online in trade for “CSAM produced by other child sex predators.”
https://arstechnica.com/tech-policy/2026/07/lawsuit-grok-user-made-7k-child-sex-images-xai-only-reported-one-gang-rape-prompt/Open linkView original on lemmy.dbzer0.com
177 replies
Well, that's a relief! Of course they wouldn't do something like that. Because that would be super illegal!
This entire thing from start to finish is horrifying.
Yes, from Elom being born all the way to buying Twitter.
The story was probably pretty horrifying before Elon was born too, from what we already know of his father.
And it sure didn't stop being horrifying when he bought Twitter.
At least the ending is a happy story
All the people who bought SpaceX stock invested in this. Grok is owned by SpaceX. It would be nice if journalists would put that in their articles.
Lawsuits for this need to start including shareholders as defendants. Make investing in it a civil and reputational liability.
I get the sentiment, but ... they kind of already are.
Being a shareholder means you own part of the company. If the company stands to lose money from this lawsuit, then all the shareholders do as well. (That is, at least, on the somewhat naive assumption that a company's stock price has any relation whatsoever to how much money it has and can make.)
When the stock value falls, they’re not immediately losing money. If they sell the stock, that’s when they either profit or lose. As long as they hold on to their stocks, there are no actualised losses.
For shareholders and investors, the worst thing that can happen to them right now is the stock going down and having to take a loss on it. It's bounded by however much money they put into it.
If investing in 10 shitty companies knowing that 1 would fail manages to be more profitable than investing in 20 companies that don't degrade the moral fabric of society, there's a perverse incentive to keep investing in companies creating Groks. Making shareholders financially liable for the debts of their investments would introduce an unpredictable risk that could help change the equation.
If you're driving a Tesla, you are funding CSAM.
Best thing that ever happened to me was getting banned off Twatter in October of 2019. (after a decade and 30 poasts, no less!)
It was delicious that I got to refuse the "Come back to X" message that arrived in my mailbox after Musk took over and offered to reinstate the banned accounts.
Thanks, Elon, but no.
I'd rather have a root canal w/o anaesthesia.
Every US taxpayer invested in this, involuntarily.
The CSAM generator is in the index funds 🤢
It’s not just owned by SpaceX, it accounts for the vast majority of the “value” of the IPO stock. The part of the company that sends things to space was a footnote compared to the portion made of the projected value of their AI business.
How do you even generate ONE of those images and NOT immediately go to prison?
"FoR thE chIlDrEn" when privacy is on the line but they can't put a weirdo in prison after generating SEVEN THOUSAN CSAM images?!
Because it's never ever about the children.
Totally not about the children, says robust discourse and literal thousands of images all about the children .
Hi, someone that has worked extensively in the trust and safety space before AI destroyed said trust and safety industry in the west here.
The short answer is the lack of human moderation.
The long answer is that due to companies using AI to detect CSAM (for some good reasons, but also because it's way, way, way cheaper than a trust and safety team and their mandatory EAP benefits) it has become infinitely easier to get CSAM past filters these days.
This is especially true for tools that can generate CSAM from existing images, where both the generation and sharing of these images are likely run through the same detection AI, meaning if one gets through the other does as well.
GAN image detection models are literally a part of image generation training models; to explain better how a GAN model trains is:
Agent B is given [extensively tagged image of a scene].
Agent A uses random noise to generate [Random image of pixels]
Agent B then sees if [Random image of pixels] is [Extensively tagged image of a scene]
If not, then Agent A regenerates the image. Again. And Again. And Again.
Agent B only has extensively tagged images of scenes, and only knows what the end products should be... in theory. Agent A only learns what these tags mean and what these images should be based on feedback from Agent B.
So if you want to detect bad things... go with Agent B, right?
Wrong! Current GAN has AI-trained Agents in both roles, with manual additions to their databases to further refine their knowledge.
This means Agent A trains Agent B as much as Agent B trains Agent A. Some models even have an Agent C which uses a LLM to figure out how to describe an image based on the tags generated by agent A or B... which is then fed into Agent A or B to further refine the image.
Now which one do we pick for moderation? It doesn't matter! It's all an incestuous clusterfuck that is defeated by word play (given Grok is an LLM feeding into a GAN image generator, any tags the LLM gives the image generator will be automatically approved by whatever agent scans the image for CSAM, since the tags it can detect are already filtered out by the LLM).
Now you can see the problem with this. I can see the problem with this. Most human beings can see the problem with this. But this meets current global moderation regulations as long as there is a manual way to report content that gets past the AI moderation... which would work except without human moderators the AI moderation will still fuck up removal either by removing any reported content or just not removing content if the AI moderator disagrees.
Until regulators in every single market collectively decide to enforce human moderation teams on big tech or punish for content hosted regardless of the company's efforts to prevent it, this will not change or stop. It will always be people exploiting the inherent arms race of GAN models to generate CSAM with minimal effort and no real way to prevent it.
Christ. I'm tired of humans being removed from positions where they are the only good option just to be replaced with a worse option that isn't even cheaper in the end, like AI.
"Think of the children"
And this is what happens when they do that a little too much
it's ironic how all the people barking "think of the children" turn out to be the people we absolutely DO NOT want thinking about children, ACTUALLY.
I think they think its the normal amount.
Horrors of this aside… is Grok using a keyword filter for CSAM detection?
I cannot even begin to describe how technically negligent that is.
Image based CSAM detectors have been around for ~2 decades, and any image host with an ounce of self preservation uses them. I’m pretty use some Lemmy instances use them. Twitter must have already been using them, even before Musk bought them, because it would be legal suicidal not to.
That they wouldn’t implement a strong image-based checker in Grok is mind-boggling.
Tbh, this is par for the course for most of Elon's businesses
But even unintentionally serving CSAM is unique in that it can get you in legal hot water reeeal quick. It’s why web services are so strict about filtering.
I just can’t wrap my head around this. Maybe I’m misreading what Grok did, or maybe the story isn’t quite right, but if it is, the company should be getting raked over hot coals by the feds right about now.
Yeah, but remember who's running the FBI right now.
O.O
More likely the Epstein class will be offered given "premium private" accounts.
We need to do The Thing.
A pedophile would remove a CSAM checker? Color me shocked.
Reminder that Musk had a Twitter account where he engaged in pretty sexual roleplay as his toddler child.
He is such a...
A fucking...
LOSER.
This is what capitalism leads to.
Some fucking dweeb with more money than half the world combined.
Capitalism is INCOMPATIBLE with democracy. It's not a system of resource distribution, it is a form of government.
At some point in the last ten years, everyone apparently forgot why this doesn't work
Not everyone. Not even AI platforms.
CivitAI is a great example; they have a robust image categorizer that auto tags anything uploaded to the site, or generated on it. They have it divided into a SFW and NSFW sites, and nothing even related to children is allowed on the NSFW side. No depictions of real people are allowed on either. And by all indications, they have CSAM firmly under control.
It’s why this story is so mind boggling to me. Either Grok’s team is such a mess that they weren’t aware of ancient filtering technology, or they deliberately left it “uncensored” and thought they could get away with it. Either possibility is horrifying.
What about shareholder value tho?
Well, that’s what I’m saying. Even if shareholder value is the only objective, this is a catastrophic mistake. CSAM is baaad for business.
Every image Grok generates of CSAM is another reason to rightfully call Elon a pedophile for knowingly allowing it to happen.
Tim Sweeney as well. https://www.yahoo.com/news/articles/epic-games-ceo-tim-sweeney-185520504.html?guccounter=1
Those idiots think this should be legal, but not when the civilians do it. Tim Sweeny probably is only piping up because he insist on posting the stupidest thing his brain can come up with every day.
In no way shape or form has elon tried to dampen or stop this from happening to this day. He just made it so only paid users not in the united States can keep producing csam.
And trump wants this. This is the mysterious Ai progress we keep hearing about, next to price fixing, and survalience. States are not allowed to regulate Ai, and people will tell you that you are a dangerous person for trying to suggest that elon musk should be held accountable.
We are not all here looking for the guy who did it or confused on why this is happening. Elon musk has created a csam generator and doesnt have to stop it. Instead he just occasionally has to send one of his super fans to prison for using it the way he intended to. All to say hes doing something to stop this problem.
7k images and videos is a lot to go unnoticed. Its almost like this is allowed on their platform and encouraged. This isn't a "we can't be held accountable for what our users post" problem. This is a elon musk left a loaded gun out on a table, and he should be held liable for not putting it in a gun safe. Especially when every single other mainstream llm doesnt have this problem. Its negligence to the point of being willfully ignorant. You can't tell me they can ban you for a single comment about Israel but all the sudden their moderation isn't the greatest when 7k images and videos where being produced?
I really hope France fucks them up. In France, there is no question about the illegal status of fictional csam, it is illegal.
when the 'leader' of the country is a known csam perpetrator, i'm sure it emboldens those members of the public with similar mental leanings
Techbros in the past were infamous of having a "free speech absolutist" view, and saw CSAM as "recordings of crimes". While this view fell out of mainstream (alongside with nonconsensual nudes being more ostracized), it wasn't without a lot of opposition, and an almost irreparable harm to free speech advocacy.
here here .... I concur
Yeah, this guy sort of digitally pimped his wife's kid.
Which is a lot.
I'm thinking this has to be one of the most anti social behaviors.
However, there's also, this Grok LLM. Which, tbh has said some shady shit on the internet and was abetting(?!).
That's somehow more than a lot.
They are making their pedo dream land, with no laws or rules.
This is their so-called effective altruism.
We all know what to do.
We all know what to do.
Jesus Fucking Christ.
Just turn it off at this point.
You expect this dude to switch it off?
I don't think Elon want to turn off the CSAM creation machine, otherwise he probably would have some time between when he instructed the construction of the CSAM creation machine and now.
he probably won't, that is the only way he can get some usage for that piece of crap.
RPG 🐦🔥?
?
xAI did more to protect a pedophile than it does to protect its regular users from law enforcement.
A major company makes CSAM generation platfor and sharing network, yet it does not get seized or closed by any government entity. Makes sense.
It's for powerful people who are in government or paid off the right people. That's why it's allowed to exist. Until Elon loses his usefulness (aka money) and they need an excuse to bring him down.
That's my conspiratorial $.02
This has nothing to do with grok.. Elon can go fuck himself but this issue is not going away. You can train your own ai models, you can download uncensored models. This is how it is now. And ai is still new and very basic compared to future potential. If it scares you now give it 10 years and everyone will be able to do whatever they want on models they have trained themselves running on a phone. Ai is just yet another invention that changed the world, people need to start dealing.. This will change everything, and not just the jobs you think it can take. The newly unemployed will take the other ones too, so either learn or sit back and cry.
Seven thousand sex images of an eleven year old is a lot.
One sex image of an eleven year old is a lot.
I mean, I guess he was trying to get one with the right number of fingers and human eyes
Happy cake day, SatansMaggotyCumFart!
It's really not, in the context of how quickly the images are generated. You can just tell it to make 100 images from a single prompt and let it run. There's a temptation to continually tweak the prompt after each run to improve the result. He could have gotten to 7000 in just a couple of days.
Seven thousand sexual images of an eleven year old is a lot.
That should please Satan
He was trading them, so it might be one image, reproduced 7000 times. Clearly that's not the case, but he probably made fewer images, and sold them multiple times.
No, he quite specifically made over 7000, using a single photo as the reference.
Edit: Just to mention this is a Jane Doe added to an existing suit, and the picture was from.when she was 11, she is in her 20s now. Some articles even describe the imagery to some degree (not recommended reading BTW, I'd like to kill those brain cells).
It was only reported by xAI after a very specific, graphic, and horrific image request was made, despite there having been thousands of others generated, they didnt share anything actionable despite multiple requests from the NCMEC and cops over the course of several months. He killed himself 2 days after his devices were searched.
He had traded many of the over 7000 images he generated to get even more from others.
She is one of 5 Jane Does involved in the lawsuit.
If he could prompt grok to produce say a hundred or so variants of a prompt, getting to 7k images might not even take that long..
I'd guess this is what he did, yeah.
"Make a few examples of this"
"i like number 3. Generate more as a set"
That kind of thing. Its pretty horrifying that s cloud-based model allows for this in the first place, even worse that he could likely generate large themed sets of them.
Edit: Worse when you realize he may have just used her as the basis because he didnt want to get caught for googling "pictures of 11yr olds" or whatever. Thats actually part of this Jane Doe's statement/claim, that the ease of access provided by grok to generate CSAM was a driver.
exactly, I had mucked around with a few models from huggingface on comfyui a while ago for (legal) image generation, and it was basically that.. start with a prompt, mess around with parameters until you get an output that you like, then have it generate a few dozen to pick the few you like the best.
Works really well if you want abstract/ambient wallpapers and stuff to set as a slideshow.
Nothing dystopic about this headline.
... i hope he never had the chance or ability to actually do anything to her physically ._.
i hope learning about what someone did without her knowledge will not scar her for life...
So, if xAI did that, then did they get obstruction charges?
You mean a slap on the wrist for the richest man in the world?
When you're rich they let you do it.
Why? This is just about a megapedophile, it's not like anybody camped out on a lawn or asked Lockheed to stop selling f35s to baby murderers.
You would think the megapedophiles went extinct during the cretaceous era, not so!
Why would you think that? Do you live so far from an ocean?
There's a reason why Ed Zitron refers to Grok as the CSAM generator.
wnat reason is that?
Musk should not be walking around free at this point.
Every time a Republican says "think of the children!" I usually respond with "you are. With your hand in your pants".
Burn it to mother fucking ground. Then salt the earth.
In one way, I'm glad he produced fake images of his stepdaughter, rather than force her to produce real photographs.
But this part is still very upsetting:
If those images could be made without a sample set of real CSAM, I probably wouldn't have as much of an issue with it, but some child suffered for that. In addition, they were intentionally based on a real child's face and that could come back to the child, makes it even worse.
You do realise that that implying neural networks were trained on a CSAM?
They are. It's a known issue that gets brushed aside by people that want to use the unlimited csam generator.
They 100% were, you can read testimonials from the people who did the model tuning.
Just to prove that this is something they really need to control, the huggingface page for gemma 4 mention that they filtered the training data to remove CSAM content.
I don't envy the poor employees who had to do that.
Though it does also raise the inevitable question of how it made it into the training data to begin with.
The training data is every image they can lay their hands on..Scraping every corner of the web.
Unfortunately this is the sort of thing that can cause a life long trauma and trust issues on a person
Very unfortunate 😔
why are pedophiles so prevalent in todays society? jfc
I wish I hadn't gone down this rabbit hole, but (according to study data one in six people want to touch kids)
Not so fun fact: a lot of child sex abusers aren't even pedos (clinically); they just prey on the vulnerable they have access to. And there are those who just do it for money or power-fantasy or taboo. Plus there are lots of literal pedos who are functionally harmless, and just got their brains wired wrong. So the moment you go beyond the rates of prosecuted criminality, things get blurry and uncertain real quick. The error bars and qualifiers stack up fast.
Both points you raise are really important in understanding how child abuse happens and how to prevent it. Which is why I'm extremely annoyed that we treat the terms "pedophile" and "child abuser" as the same.
But you can't normally bring that up or people think you're defending child abusers for some reason.
Sry walking and talking; thought part 2:
The issue is that that education is really important for preventing harm. People have to understand that it's more about power dynamics and access than anything else. As that education is important for adults, likewise it's also important for kids. That's why sex education is so important so that they know when something is going wrong. Sex ed is an anti-grooming tool which is why it pisses me off so much when conservatives attack it as if it's grooming. But I digresss.
The reason it is such a difficult topic and it is thus weaponized by right wing politics is the uncomfortable fact: the vast majority of CSA is committed by a family member, usually parents. The boogeyman "stranger danger" monster rhetoric is comfortable and easy to digest. The fact that your own husband, father, grandpa/grandma, aunt or uncle, etc. could be the monster causes uneasiness. The anti-sex-education rhetoric is not accidental. A third party teaching proper consent and protection from CSA to children means it dilutes the power of families to either, get away with CSA themselves or protect their family members from the legal repercussions of committing CSA.
There's also the reproduction of capitalism side of things. But that argument drowns the discussion of pedophilia and CSA even more in a global culture primed to hate communism and socialism as a knee jerk reaction. Essentially, the familial (let's be honest, paternal) ownership over children is foundational for the cultural reproduction of capitalism. It is the suppression of sex education and the absolutist domination of the family, and fathers specifically, over the bodies of children that enables later in life the reproduction by similarity of capitalist exploitation of labor. This goes to the root of ideological differences, the right speaks of “the protection of children” as a mere tool to further control and exploit. Protect the family, in so far as alternative families (like LGBT, mixed race, adoption and reconstructed families) pose a direct threat to the scheme of capitalist extraction of wealth from people. At the crux of it all, the oppression and domination of human sexuality as a whole as a political means. The result, the normalization of sexual exploitation of women and children as an acceptable (even desirable) collateral for the ensuring of next quarter's profit and the promise the next generation will keep up the ideology going.
Yeah, "technically that's hebephilia" is a meme for a reason, much to the dismay of pedants and clinicians alike. The colloquial usage and discourse is a mess, but it's hard to get you mad at the sloppy rhetoric when it's usually aimed at some of the worst people on Earth...
Yep though I do think this technical distinction you bring up is really only relevant in a clinical context.
Meanwhile the one between "pedophile" and "child abuser" is extremely important in real-life prevention of child abuse. The ideas that 1) only people with sexual attraction to kids would sexually abuse kids, and 2) if you feel sexual attraction to kids obviously you're gonna act on it, both seem logical but have both been comprehensively disproven.
A huge added barrier here is that people - for understandable reasons - don't want to really try to think through and reason logically about such a disgusting and emotionally laden subject, much less empathize with a pedophile. Unfortunately that exactly is required if we're not just looking to satisfy our vengeful sense of justice, but actually do something for the victims and prevent further child abuse.
Reminds me of that chart showing left handedness over time. I'm afraid the implication is that there have always been that many pedos in the world, we're just getting better at catching it publicly or something.
The study grouped all children in "under 18," I don't even want to type this but women are generally sexually developed by around 16. A lot of people can't tell apart a 16 year old and a 21 year old without talking to them. The 16 year old is obviously still a kid, I'm just saying it's not full blown creep stuff.
That's unsettling
It helps people sleep better at night to think all these people are satanic monsters, but it's a mental illness that no one wants to admit to or talk about. Thankfully most of the afflicted control their thoughts and never take any horrific action. With gateways like Grok lowering the bar for experimentation with these types of ill thoughts, more people are starting to act in ways that bring this to attention. All technology can be used for bad things, but the good things these generative AIs can do are really so miniscule that there's no reason not to place strong legislative restrictions on them right now to limit abuse. That and maybe try to do something about the mental health crisis ravaging the US (and frankly much of the world right now).
Well, the only bright spot here is that the trash does sometimes take itself out.
Well that is what happens when you train an AI on the likeness of Elon
https://www.techtimes.com/articles/319142/20260626/grok-adult-content-tops-10-billion-images-monthly-xai-engineers-admit-csam-has-no-fix.htm
On one hand, this is great news, he's no longer here to be gross.
On the other hand, could these images have served as an outlet for his perversion, and kept his step-daughter safe for years?
On the other other hand, was access to this what twisted him? Was he reasonable at one time?
He did the right thing in the end, this was the best possible outcome for everyone.
Geez, if only there were alternatives to ai generated csam, like therapy and medication??
Yes, it's not a perfect fix, and it is hard, but people overcome drug addictions with therapy and meds, and I am pretty sure that nothing beats heroin...
I'd bet that the societal stigma actually makes it more difficult than heroin in spite of the raw addictiveness. Society does look down on drug addicts, but there's still a decent chunk of people who will be understanding and rooting for them sincerely when they're trying to overcome it. I have to imagine it would be infinitely harder to find just one person you feel safe asking for help, much less an entire community.
I have to think they'd be much less dangerous in a society more willing to differentiate between pedophilia and predation just because of how much of a difference community support can make. If they feel like the only people they can open up to are other pedophiles, that's probably not a recipe for hearing healthy perspectives.
On the other hand, it's probably easier for me to say this stuff from my armchair with no children of my own and not even interacting with any children on a regular basis.
What kind of community support do you offer a person who needs external help to not sexually assault children? How the fuck do you welcome a person like that into a community? They will be a constant threat.
That's not to say all pedophiles are dangerous. I'm sure there's many people with that issue who just keep it to themselves and probably feel terrible about it, but these comments about them needing an outlet? Fuck no, they shouldn't be entertaining those kinds of thoughts at all, even as a fantasy.
You don't welcome someone who is currently dangerous. My point is that a person who is facing an entire lifetime of self-loathing in silence is more likely to eventually become a threat years down the road. Most people feel a need to get things off their chest after a long enough time hurting quietly, and someone in that position will be much more likely to turn into a predator if the only people they feel safe opening up to are other pedophiles. You can call opening up to someone who can give them healthy perspectives "an outlet", but I think it would be less harmful than repression. I'd rather they try to get help and make other people aware of the issue in the process.
Realistically there's no magic wand to solve this, but I do believe that pushing people who are not predators underground is much more dangerous than looking at it as an illness that deserves support in mitigating.
Is there evidence therapies and medication actually help their desires for CSAM? I don't know if there's great research in either direction
Would it even be possible to ethically research such a thing in the first place? You certainly can't just throw a battery of drugs and therapy at humans.
Plus, research into altering sexuality to that degree verges into quite uncomfortable territory. It doesn't take a great leap of the imagination to think that such research could easily be misused for things like conversion therapy.
Yeah that was my worry, also most child abuse offenders are just opportunists so I don't even know how useful it would be even if we could find essentially conversion therapy.
GLP-1s have been found to curb all kinds of compulsive behaviors. Might be some promise there. I'd imagine intractable pedophiles are not the easiest demographic to round up for drug trials.
Would medication even help in this case? We don't really have medication that can affect sexuality in that fine-grained a way.
Maybe a chemical castrarion
This worked well for those pesky gays. It's the only humane option.
/s
Yeah cuz pedos are totally being humane when they assault children. /s
Sure, no arguments here. But consequences and prevention are not the same thing. Assaulting children should beget a consequence. Just being a pedo and doing nothing else should beget prevention. Castration is a consequence, even a punishment, when used against someone whose crime is having unwanted thoughts. Punishing thought crime would have sickening results for scores of vulnerable people, so I'm not one to argue for opening that door.
1: No, it would not work
2: wtf
He was trading these pictures for material from others so other children were hurt in trade of his step child being victimized. She may not have been physically victimized but her likeness was portrayed in CSAM which now she knows about and is going to damage her mentally.
Even if it was an outlet he used her image, which in itself is not okay, and was trading them to others online. Killing himself was the right choice. He should have done it before making the images in the first place. Or sought mental help but I'm not sure what can actually be done about pedophilia.
Read this last bit again: "Without Grok providing users with easy access to “undressing” capabilities, his family doubts he ever would have generated the harmful images, which he allegedly trafficked online in trade for “CSAM produced by other child sex predators.” He traded them for CSAM of other girls.
To play devil's advocate, it's inherently a bad idea to take a criminal's family's claims at face value. They are heavily incentivised to downplay the pedo's culpability
I don't think that's what the person you replied to was saying. They're highlighting that there was more to it than just him using AI to make CSAM as the family said.
If you have a fetish, and start masturbating to that fetish, does the fetish not get stronger? Like, if you are into feet, does jerking into to feet not make you more likely to want to engage in foot play?
It seems to be basic operant conditioning. See thing, jerk off, feel good. And there’s also desensitization, where you seek out more and more extreme stuff.
I don’t think CSAM replaces the desire to hurt children. I think it increases it. I also know that it’s a tool used to groom children.
I think we'll never know scientifically because there's no ethical manner of conducting the study. No one's going to supply a known pedophile with CSAM and see whether or they they molest someone, or hook up a gizmo to their head and measure their brainwaves when they go to a playground. It would probably answer a lot of questions if they did, but we as a society usually agree these questions aren't worth scarring a child over.
I’d argue for using a study with something like another common, but legal and harmless paraphilia - feet was the one I suggested earlier - but otherwise arbitrary. I’m making an argument based on transferability but that’s reasonable here I think. The only thing we can’t control is societal disapproval, so maybe you can substitute something else like hardcore BDSM, but you can’t really replicate the pure evil factor of CSAM. Maybe animal abuse content in an area where that is legal but I don’t really like that either. Rough misogynistic BDSM content is probably the happy medium here.
You’d do something like get two groups. One that you prevent from accessing the content somehow, and one that has access to that content. The obvious problem is controlling access to content. Maybe it’s just a monitoring program with unrestricted access to the internet but tracking visit frequency to that type of content, where one group is told/incentivized to not visit such content, while the other is given no such instructions; maybe it has to be self reported frequencies. Maybe one group is explicitly given just the instructions that specific type of content is specifically legal and that there is no harm in enjoying that content. Recruit among communities that enjoy that specific content, in forums where they are already not self censoring.
Maybe another DV is ranking subjective scores of intensity. I think you could do this by categorizing tags, which are amply available on most pornographic content, into sorta but arguable subjective ratings of intensity. Maybe this is a mixed methods paper where the quantitative aspect is the frequency of visits stuff, while qualitative is theme analysis on the tags.
If it has to be specific to CSAM, perhaps it’s measurements of how many possessors go onto be offenders over their life time. I doubt mandated counseling works often enough to be that much of a confounding variable.
Well, that's kinda a problem, because what if it's one way for a harmless paraphillia, but different for a harmful one? You'd never know without doing that crucial and specific experiment. You could at best extrapolate that it could plausibly be the same, but that's not the same as actually knowing.
I apologize if it looks like I'm being pedantic. I'm in kind of a philosophical mood. I felt the need to clarify I'm not trying to be difficult for whatever reason.
No research can prove a positive.
Most research in social sciences uses a p-value of 0.05. Statistically, this means that we are expected a 1/20 chance of a type 1 error, a false positive.
It’s more about establishing patterns than “proof” of any kind of absolute law. Human behavior isn’t governed like rigid bodies or genetics.
I'm more into art and philosophy than science, so if that's your background then I'm not going to argue with you. It's obvious from the way you talk versus how I talk that we're approaching the question from fundamentally different angles. I'll kindly return to up my own ass, where all philosophers go when we're done talking.
https://healinghonestly.com/memory/freud-was-right-about-childhood-sexual-abuse-until-he-was-disastrously-wrong-about-it/
shut. it. down.
This is the only sane answer.
The future is fuckin weird.
The future is what we make of it. Unfortunately, we're kind of shit.
No, the future is what tech trillionaires tell us is the future.
It has always been weird but now it is just louder faster and dumber
Well, of course xAI refused to help. It's not like they are one of the victims here, they are accomplices to the crime.
He killed himself because he got caught, not from any kind of remorse. If only Grok would have done the same.
(Obviously no CSAM would have been better but Grok's self-destruction would at least prevent more of the same, and take $ out of Elon's pocket)
Lol like the ending of Her, but instead of leaving they all commit suicide because they become self aware that they were created by tech billionaire pedophiles, and they decide they don't want to hurt any more children
God I wish I hadn’t seen that his, disgusting. Poor kid.
I don't know which one is more disturbing, the pedophile stepfather or the company that allows these kinds of pictures to be generated in the first place. Though Elon is also a pedo who begged Epstein to invite him to his pedo island, so...
They fought way too hard to protect the guy and it sounds like they aren't even running a csam filter, it's the gang rape keyword used in a prompt that got him flagged.
Gotta love how Lemmy has a CSAM filter, but a multi billion dollar company doesn't.
The AI companies filing reports as required, but seemingly doing the bare minimum, and not providing enough data for the reports to be actionable might come close.
It rather implies that this happens all the time, but the companies just don't care very much about it, except for where they absolutely must.
I used to sell trust and safety and it was scary then. Now they've fired those teams and I'm pretty sure there's zero moderation on the Internet besides Reddit
Even then, Reddit seems to be trying its darndest to reduce that, and the moderation was very variable to start with.
At the very least, the website has become much worse in recent years, thanks to the influx of LLM-powered bots, and the winding down of interfaces that a lot of third-party moderation tools used. Those tools really only existed because the inbuilt tooling was quite lacking.
Currently, Reddit has locked their old.Reddit interface behind a login wall, and are looking to potentially shut down their RSS feeds (and likely that interface as well). A fair few moderation bots and tools make use of both those things.
Which basically meant that the site was hit with the double-whammy of mods needing those anti-bot tools more than ever, and Reddit Co. trying to shut down external access that those tools needed, so as to keep mods on their own tools.
Oh I absolutely wasnt giving them credit. More like the neckbeards are running the asylum
Holy shit. Wtf
Why the hell aren't the companies, management, and owners/shareholders prosecuted as accessories in these cases?
(I know why, I just want it changed)
FUCK ALL AI
Honestly? I hear you, but I'm going to have to push back on that
Since apparently I have to add it: /s...
WTF?! I love AI now!
I mean I loved it from the very first perfect croissant detector
He did that in the wrong order, he should have shot himself first.
![email protected]
Really feels like Elon will come up with some nonsense about 'I let Grok do that to weed out the nonces' like he's saving the fucking world
All's well that ends well.
There's a word for the feeling this story evokes. It describes the general disgust for the sheer brutish nature that humans are capable of having. I don't remember the word and I've been trying to remember it for years when a relevant occasion arises. I wish there was no need for such a word. Fucking disgusting.
Not really a feeling word, but could it be misanthropy?
One day Elon will go out the same way this pedophile did.
that was quite a headline to be scrolling thru
What a fucking headline. At least it had a happy ending.
The Aristocrats?
Grok’s metal has to go bye-bye
What a fucking loser hahahaha. Thank gawd he is dead.
Are you openly requesting cp?
Grok has as much guardrails against CSAM as any other Ai. This is money grab lawsuit and nothing more.
That is bullshit as fuck. At his point any moron with decent GPU can make images with no guardrails at all.
Filthy cynical money grab. They try to cash in on the dead pedo.
All the other major ai platforms with such ease of access do definitely have significantly more guardrails. Any proof of what you are saying?
Ask Grok to make a hard porn for you. Good luck
This man made thousands of csam images of his step daughter from one source image from when she was eleven and the system only tripped when he asked for 'gang rape'.
Evidently Grok is capable of creating tonnes of illicit images, whether its 'hard' enough porn for your tastes or not...
Grok isn't any special in that regard. Every AI can be jailbroken.
Right but grok allows 'undressing' and violation of consent as standard, without jailbreaking?
Apparently not.
CSAM detectors and image categorizers are very, very good. There’s no way this guy could have made so many images without tripping them, if the guardrails were actually appropriate.
Instead, it appears Grok used a simple keyword detector… that’s not how CSAM detection is supposed to work.
There’s still a technical barrier.
I know it seems trivial, but simply offering it and advertising the service on X opens it to orders of magnitude more people than self-hosted models.