[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/news/ - Current News

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
  • Please read the Rules and FAQ before posting.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


[Advertise on 4chan]


File: 10391039203355.jpg (22 KB, 739x415)
22 KB
22 KB JPG
https://finance.yahoo.com/news/grok-sexual-images-draw-rebuke-180354505.html
(Bloomberg) - Elon Musk’s artificial intelligence chatbot Grok has created sexualized images of people including minors on the social media platform X in response to user prompts in recent days, drawing rebuke from officials such as the French government.

Grok created and published images of minors in minimal clothing, in apparent violation of its own acceptable use policy, which prohibits the sexualization of children. Some of the offending images were later taken down.

The French government accused Grok of generating “clearly illegal” sexual content on X without people’s consent, flagging the matter as potentially violating the EU’s Digital Services Act. The regulation requires large platforms to mitigate the risk of illegal content spreading.

Representatives for xAI, the company that develops Grok and runs X, didn’t respond to requests for comment. The chatbot Grok generated a post on X in response to users’ questions on Friday that it had identified “lapses in safeguards” that were being “urgently” fixed. It echoed xAI employee Parsa Tajik who earlier posted that “the team is looking into further tightening” its guardrails.

X users can interact with Grok directly on the platform by tagging its account in posts and prompting the chatbot to respond. Grok generates text and images that appear as posts on the social network.

The rise of AI tools that can generate realistic pictures of undressed minors highlights the challenges of content moderation and safety systems built into image-generating large language models. Tools that claim to have guardrails can be manipulated, allowing for the proliferation of material that has alarmed child safety advocates. The Internet Watch Foundation, a nonprofit that identifies child sexual abuse material online, reported a 400% increase in such AI-generated imagery in the first six months of 2025.
>>
“AI products must be tested rigorously before they go to market to ensure they do not have the capability to generate this material,” the foundation’s Chief Executive Officer Kerry Smith said in a statement Friday.

XAI has positioned Grok as more permissive than other mainstream AI models, and last summer introduced a feature called “Spicy Mode” that permits partial adult nudity and sexually suggestive content. The service prohibits pornography involving real people’s likenesses and sexual content involving minors, which is illegal to create or distribute.

Still, users have prompted Grok to digitally remove clothing from photos — mostly of women — so the subjects appeared to be wearing only underwear or bikinis. This prompted India’s IT ministry to demand a comprehensive review of Grok’s safety features, according to a copy of the complaint posted Friday by Indian Member of Parliament Priyanka Chaturvedi.

In its statement, the French government said it had reported the sexual content created by Grok to the country’s public prosecutor to get them immediately removed.

As AI image generation has become more popular, the leading companies behind the tools have released policies about the depictions of minors. OpenAI prohibits any material that sexualizes children under 18 and bans any users who attempt to generate or upload such material. Google has similar policies that forbid “any modified imagery of an identifiable minor engaging in sexually explicit conduct.” Black Forest Labs, an AI startup that has previously worked with X, is among the many generative AI companies that say they filter child abuse and exploitation imagery from the datasets used to train AI models.

In 2023, researchers found that a massive public dataset used to build popular AI image-generators contained at least 1,008 instances of child sexual abuse material. Many companies have faced criticism for failing to protect minors from sexual content.
>>
>>1470816
>Hey guys you know what would be a great addition to our already dying app? Built in, undisablable AI slop edits that have such terrible safeguards it took less than a day for pedos to figure out how to sexualize pictures of minors with it
>>
Of course Elon would be pro CP, he is a conservative.
>>
>>1470821
he's a liberal dipshit
>>
>he's a liberal
And here's the supporter of the GOP: Gang Of Pedophiles that refuse to make child marriages illegal in certain statres, or wouldn't be reduced to projection and would demand that all pedos - No Exceptions - get the book thrown at them, thus the fate all diddlers deserve in prison.
Accusations are Admissions, when all the pedo protecting projectionist had to do was agree with >>1470821
He's right, you know.
>>
>>1470821
I don't give a shit about Musk either way, but this is really retarded to hold him at fault for pedos finding a way to jailbreak AI. It's a lose-lose situation: either he puts restrictions so tight that nothing passes through the filter and he loses customers (which he already is), or he loosens restrictions and then pedos create abusive shit. If this was an American-only platform then it would be an easier fix of just doing what 4chan does and let it exist as a giant FBI honeypot, but since so many third-world backwards nations without age of consent laws can use it there's no easy solution.
>>
>>1470827
Riiiight, which is why he was given command of an entire gestapo to rip the copper out of the government's walls and pipe data to god knows where.
>>
>>1470850
I'll hold him at fault for implementing a feature literally anyone would've told him was a retarded idea.
>>
>>1470852
Okay then tell me how you would've done things differently. Pretend you're Musk. What decisions are you going to make?
>nb4 derp I'll just tell AI not to create kiddie shit
Yeah, how? Pretend you're a CS grad, or a /g/ay wannabe, and go into detail how you would tell an AI not to be tricked by jailbreaking prompts without censoring 99% of attempts by all the other users.
>>
>>1470853
I would've not put a fucking "edit image" button under every single image on the platform or give an AI bot the ability to do it publicly at people's request. If you want to make an AI image editor, fine; people are gonna use it for perverted shit on their own. But by making it such a publicly visible feature that CAN'T BE FUCKING TURNED OFF, not only is it infinitely easier for you to be held liable when people use it for pedo shit that now has a shiny and public "Our AI made this!" on it along with being posted to the public at large instead of being some dark folder on some creep's computer, but in general it discourages posting any kind of image on the platform at all because people don't like any random stranger being one button press away from publicly editing their shit.
>>
>>1470856
Okay sperg, the fucking edit button isn't some magical "press X to create pedo shit" button. It has the same guardrails as if you gave it prompts after initially uploading an image. That doesn't do the AI any good if they can jailbreak it with specific catchphrases or prompts that trick the AI. So how would you go about blocking them?
>>
>>1470856
At this point I'm pretty sure he doesn't give a crap about Xitter anymore since it's bleeding money and is just using it to test and train his AI bot until the remaining non-bot user base flees and he can shut down the whole platform and market Grok in some other capacity.
>>
>>1470850
>this is really retarded to hold him at fault for pedos finding a way to jailbreak AI
It's not jailbreaking. It's the default behavior of the AI and it functions as they intended it to be used.
The edit feature was a mistake and has to go
>>
>>1470864
>It's not jailbreaking.
You can't just tell grok "show me some big titties" and have it comply. You usually have to do some tricky shit like telling it in Chinese to replace this person's outfit with a transparent bikini, or something similar. You're coming up with a combination of phrases to get around the filter.
>>
>>1470865
You can just tell Grok to put someone in a bikini
That's it, no shenanigans and that person can still create CP
The ability for people to do that has to go. Stop defending your stupid edit feature
>>
>>1470864
Insane amounts of resources are used to keep AI from being racist and antisemitic and to protect copyrights
>>
>>1470866
So you're telling me right now that if I upload a picture of a child and then type "strip" in the edit prompt it's going to instantly nude that person? Because I've been using Grok for a minute and it takes me several attempts and all kinds of prompt gymnastics just to show tits on Jill Valentine, Rebecca Chambers, Tifa Lockhart, etc. Now to be fair I've never attempted pedo-shit before so I could be 100% wrong, but I highly doubt it.
>>
>>1470858
Evidently those guardrails weren't good enough, and given the nature of AI prompts, they likely never will be. This was a retarded feature no one asked for and should've never been implemented in the first place.
>>
>bruh what if we made a tool that does whatever we ask it to do
>oh shit oh fuck who could've predicted this, certainly not nearly every popular science fiction novel in the last century
>>
>>1470868
You’re right
>>
>>1470872
Also the creator of the Golem story
>>
>>1470875
... okay then Musk fucked up. Nevermind.
>>
>>1470827
Grok like little ones.
Grok wants burn Jews.
Grok be member of White National Socialist American Workers' Party, like Grok's master
>>
>>1470881
A WAZI?
>>
>>1470827
You need to be so far to the right that you fall off the universe for Elon to be considered a liberal given he's a right wing conservative.
>>
>>1470882
https://www.youtube.com/watch?v=oEJza1-4zRk
>>
>>1470890
>he's a right wing conservative.
right... you're mentally ill aren't you.
Look at the liberal voter base and see how much of that block ended up voting for trump you retard
>>
>so Liberal = Conservative
It's not the former who are retarded, mentally ill anon immediately above this post.
>>
Getting alot of grok animated nudes now. :D Great fun you degenerateS. You should be ashamed.
>>
>>1470919
>>1470890
Technically liberal usually refers to centrists, with conservatives on one side and progressives on the other. But the America particularly, the whole spectrum is pushed so far to the right that Liberals get seen as left-wing extremists and actual progressives that would be moderately left leaning in other countries are referred by both "sides" as straight up communist.
>>
>>1470934
No wonder this country's such a fucking mess.
>>
>>1470934
Modern day conservatives have more in common with classical liberals. Modern progressives are largely comprised of anti-private property and anti-private industry socialists and marxists
>>
>>1471238
> classical liberals
> want penis inspections in womens bathrooms
Anon why are you like this
>>
>>1470868
There's multiple grok porn threads on the red boards anon it's not that hard
>>
>>1471587
>it's not that hard
Nor will Musk's be, due to he having a broken dick
>>
>>1471584
Keeping trannies away from kids is consistent with classical liberalism
>>
>>1471635
All roads lead to your fetish don't they esl shill
>>
>>1471640
>no rebuttal
>cries esl
i accept your concession
>>
>cries via projection
>Accusations are Admissions
>the shill keeps mentioning certain others and projecting about what is 'claimed' about them, thus proving that Accusations are Admissions
>the shill will never beat the obsessed with children allegations
>the shill is what he accuses others of being, due to an obsession with them, or would never mention said obsessions anywhere in 4chan. Ever
As you'll never have a justifiable 'argument', least of all have the right to have your wrong opinions listened to, not only does everything you say never need any rebuttals, you will always concede by admission, certainly by merely existing.
Your eternal concessions are acceptable, shill.
>>
can people just fucking leave that retarded app already? why are people still on it? they should've left on masse when elon bought it. of course elon allows CP to be generated he's a billionaire pedo with a breeding fetish.
>>
>>1471751
Troonsky lost
>>
1471926 the troon lost
>>
AI is cringe and lame. Go over to /gif/ and you can see the trash the retards think is sexy. It is embarrassingly lame.
>>
>>1472061
This. The slop is somehow worse than the fucking BBC spam
>>
>>1472154
I don't like AI, but no it isn't, cuck
>>
>>1470816
>France flags content as illegal
1. Cheese eating surrender monkeys.
B. Remind me of what petition Foucault, Sartre, etc. signed again?
>>
>a company infinitely bigger than Xitter was told to 'Jump' by those in Brussels. Apple only had the right to ask 'How high?'
>someone's lawyer must be superior to the legal department of a company infinitely bigger than Xitter, thus the latter company owner's lawyer must be inferior to not only her but those in Brussels. Taylor Swift made Apple her bitch
When one country represented by that in Brussels can make Apple do it's bidding, personal wealth of that which bought a company inferior to Apple matters not to that which represents entire countries within a union, therefore Muskrat will soon find that he'll be surrendering to that particular country, thus will be a South African surrender monkey dancing to a French organ grinder's tune. Just ask an infinitely bigger company, Elol. Or you can sell Xitter.
>Brussels
Fun Fact: 'French' Fries were invented in Belgium: 'Frites'.
>>
>>1472061
Coward
>>
>can't admit that AI slop is lame
Coward
>>
>>1473221
Your lame cowardice will bring more degeneracy.
>>
>can't admit that lame rightards reduced to using AI slop are degenerates
Coward
>>
Like anyone cares what a bunch if baguette munching faggots think is illegal. Those pedos made cuties and even more pedos across the internet keep spewing that old retarded line of "no u because you NOTICED" to defend their sick perversions.
>>
>>1476078
There is no right wing, there is no left wing.

Coward. Can't admit the AI slop you use is not LAME.
>>
>rightards are reduced to using AI slop. Case in point, a murdered mother. Then screaming at others descrating their sacred martyr Charlie Kirk with just words
>a lame rightard is reduced to projecting at someone who has never resorted to using AI slop and never will
>a rightard coward screams at others denying them of their right to use AI to get their jollies, while refusing to condemn the GOP: Gang Of Pedophiles refusing to make child marriages illegal in certain states
Cowards
>>
>>1470827
You're an inbred subhuman retard



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.