Three teenagers from the US state of Tennessee sued xAI, Elon Musk’s artificial intelligence company, accusing the Grok chatbot of generating pornographic images of them from real photographs.
The collective lawsuit, which could involve more than a thousand underage victims, is directly related to the proliferation, at New Year’s Eve, of hyper-realistic montages (deepfake, in English) of naked women and children, which led to the opening of investigations in several countries and in the state of California.
The complaint, filed in a federal court in San José (California), cites the case of an author — already arrested — who used Grok to transform ordinary photographs of young girls, taken on social networks or in school albums, into hyperrealistic sexualized images.
These montages later circulated on X, Discord and Telegram, later migrating to dark webserving as a bargaining chip for other child pornography contentreports the complaint, according to the lawyers representing the three teenagers.
“Seeing my daughter having a panic attack when she realized that these images had been created and disseminated with no hope of deleting them was devastating,” said the mother of one of them, quoted in a statement released by the lawyers.
xAI “deliberately designed Grok to produce sexually explicit content for profit”, without implementing protective measures used by other major players in the AI sector against child pornography, the lawyers denounce.
The complaint is based, in particular, on two North American federal laws: and Peoplewhich allows victims of child pornography to obtain compensation, and the law on the protection of victims of human trafficking.
The complainants demand compensation and ask that xAI be prevented from allowing this type of assembly.
Platforms are largely exempt from liability in the US for content posted by users. But “without xAI, this illegal content would never have existed”, he argues Annika K. Martinlawyer at Lieff Cabraser.
3 million sexualized images in 11 days
According to a study by the Center for Countering Digital Hate (CCDH), Grok will have generated around three million sexualized images in just eleven days at the end of 2025, of which 23,000 represented minors.
Faced with the wave of outrage, xAI restricted, in mid-January, the generation of images with Grok only to its paying subscribers and guaranteed to block the generation of sexualized images “in jurisdictions where this is illegal”.
Musk questions freedom of expression
Elon Muskwho speaks daily on X, criticizes government regulations, accused of wanting to “suppress freedom of expression”.
In mid-January, platform X announced measures to prevent have your Grok AI chatbot turn photographs of real people into images of a sexual nature, “where it is illegal”after criticism around the world.
Apparently, then, in moral code of Elon Musk’s companiesit is okay to generate, without consent, sexualized images of women and children — as long as it is in a country where this practice is not illegal.
