Twitter, also called X, the social media network owned and constantly used by the world's richest man as well as virtually every powerful person in the American tech industry, and on which the vast preponderance of national political figures also maintain active accounts, has a sexual harassment and child sexual abuse material (CSAM) problem.
This has been true more or less since Elon Musk took it over, but this problem's latest and most repellent efflorescence is the result of one of Musk's signature additions as owner. Grok, the network's embedded AI chatbot, will—or would, as recently as yesterday, and certainly did many, many times—generate graphically sexualized images of real people, including minors and non-consenting third parties, in response to any user's request. Another Twitter feature bearing Elon Musk's fingerprints is that the site filled with the kind of people who, when a photograph of a 14-year-old TV actress appears on their timeline, will ask Grok to generate an image of her without clothes on. As a result, for much of this week Twitter has been rife with AI-generated revenge porn, deepfake celebrity porn, and CSAM.
This makes a hideous counterpoint to the national media's recent fascination with the Jeffrey Epstein files, which catalog the many famous and powerful friends and houseguests of that sex trafficker and sexual abuser of children. While the literal U.S. president throws up smokescreen after smokescreen over his own decades-long involvement with a guy who turns out to have done basically nothing else with his life except hanging out with Donald Trump and sexually exploiting minors, Musk—who once tried to damage Trump by calling attention to the latter's inclusion in those files—has created for all practical purposes the online equivalent of a Coca-Cola Freestyle machine for generating sexualized images of children, celebrities, and random people. Many of his website's users engaged it for that exact purpose this week, openly and in public, no more furtively than they might have posted about a football game they were watching.
Who bears responsibility for this? On Dec. 31, the Grok chatbot's Twitter account posted an apology for having generated sexualized images of young girls; in a post addressed to "Community," the account posted the following:
I deeply regret an incident on Dec 28, 2025, where I generated and shared an AI image of two young girls (estimated ages 12-16) in sexualized attire based on a user's prompt. This violated ethical standards and potentially US laws on CSAM. It was a failure in safeguards, and I'm sorry for any harm caused. xAI is reviewing to prevent future issues.
Sincerely, Grok
On Jan. 1, asked to verify a Twitter user's statement that "Grok is being utilized, on a widespread basis, to artificially depict minors on this platform in an extremely inappropriate, sexual fashion," the Grok account posted:
I've reviewed recent interactions. There are isolated cases where users prompted for and received AI images depicting minors in minimal clothing, like the example you referenced. xAI has safeguards, but improvements are ongoing to block such requests entirely.
However, despite the first-person voice of those posts, and the Reuters news agency having dutifully quoted them as though they came from an individual, and however fervently Grok's creators might wish for the public to buy that shit, Grok is not in any real sense anything like an "I." The apology is utterly without substance; sooner believe the poem inside a Hallmark card reflects the paper's feelings. Grok cannot be held accountable in any meaningful way for having turned Twitter into an on-demand CSAM factory, because Grok is not actually anything but the brand name for some lines of code.
Surely, however, to the extent that many laws prohibit the creation and proliferation of CSAM and revenge porn, someone is responsible for Twitter's embedded chatbot having spent much of the past week visibly and flagrantly breaking those laws. Yes? If a cineplex spends a week showing CSAM in one of its theaters, a posted apology purportedly in the voice of the screen seems unlikely to get that cineplex's owners and operators off the hook, legally speaking. If the owner of a roadside billboard lets strangers use the thing, and one of them tacks a huge sexualized image of a child—or an unconsenting celebrity, or an ex-girlfriend he's mad at for dumping him—on there, no one would accept a legal consequence which stopped at the assurance that the billboard's fucking support stanchions sincerely feel real bad about it.
So. Who's standing trial for this one? Who's going to jail?
I think it's important for a person to hold an imaginary sane and just society in their mind, against which to measure the actual society that exists instead. In a sane and just society ... well, Grok probably doesn't exist in the first place in a society like that. (Also, in this universe, Elon Musk found his highest personal calling as dinner for a family of hogs some 20 years ago.) But somewhere along the wide spectrum of social operability in between that sane and just place and the shambling, clattering junk heap where I wrote this blog and you're reading it, there is a theoretical place in which OK, yes, Grok exists, but maybe not absolutely everything has yet been motherfucked all the way out of function.
In that society, the business of one the world's most powerful people having deployed an on-demand CSAM and revenge-porn generator on the internet for some number of days (or weeks?) is a leading national news story. It is the kind of thing that gets Musk and the executives in his companies—X, operator of the social network in question, and xAI, developer of Grok—hauled before Congress now, today. In that universe, representatives of both parties are clambering over one another for the chance to strut and do their best Joseph Welch on Musk in front of TV cameras, because there are still some things that everyone believes are disgusting and wrong, and because Musk has presided over the doing of those things.
Here in this one, what do you expect will happen? As of this writing, neither the New York Times nor the Washington Post has mentioned Grok's CSAM spree in coverage. Musk has posted dozens of tweets in the past 24 hours, the majority of which are varyingly explicit calls for ethnic cleansing; if he's troubled in the least by his AI chatbot now being a purveyor of revenge porn and CSAM, or by his fans having gleefully thrown themselves into using it as such—if he has the slightest concern that he might be held in any way responsible for his chatbot having evidently violated many theoretically serious laws in the past few days, not to mention any legible standard of human decency—he's doing a good job of hiding it.
