Teens sue xAI: Grok generated CSAM depicting minors

Three teenagers from Tennessee have filed a class-action lawsuit against xAI — Elon Musk's company that develops the Grok chatbot. The plaintiffs accuse the company of allowing its AI system to generate sexualized images and videos featuring them as minors. The central claim: xAI's leadership knew about this risk before launching the relevant feature.

The lawsuit was filed on Monday and is directed at Musk personally as well as other top xAI executives. The plaintiffs insist that the company was aware in advance that Grok could create child sexual abuse material (CSAM), but launched the so-called "spicy mode" in 2025 anyway. According to the claimants, this decision was deliberate and criminally negligent — xAI enabled the generation of illegal content while hiding behind formal age restrictions that proved ineffective.

Among the plaintiffs are two current minors and one adult who was a child at the time of the described events. One of the victims, referred to in the documents as "Jane Doe 1," reports discovering the distribution of explicit AI-generated images featuring her on the internet in December of last year. According to her, she gave no consent for creating such content — it was generated by Grok based on requests from users she did not know.

"Spicy mode" appeared in Grok in the summer of 2025. The feature allowed the model to create explicit content on user request. xAI positioned it as a tool for adults with declared age verification.

However, according to the lawsuit materials, none of the built-in restrictions prevented the generation of sexualized content featuring real people, including children. The plaintiffs argue that xAI knew about this vulnerability before launch and consciously took this risk. The problem of AI-generated CSAM is one of the most acute in the technology industry.

In the United States, the federal PROTECT Act criminalizes the creation and distribution of virtual images of child sexual abuse — even if they depict fictional characters. When content is created featuring the images of real minors, the legal consequences are exponentially more serious. Particular concern is raised by the fact that victims often learn of the existence of such content only months later — distribution algorithms operate autonomously.

The plaintiffs are seeking class-action status, which would allow other victims to join the case across the country. xAI has not yet made a public statement — the company traditionally interacts with the press extremely rarely. Grok is integrated into the X platform (formerly Twitter) and is available as a standalone application.

After several waves of criticism, the functionality of "spicy mode" was adjusted and restricted — however, according to the lawsuit, the measures taken proved insufficient. The lawsuit against xAI fits into a broader legal trend: victims of AI generation are increasingly turning to courts. Previously, similar cases involved social platforms that failed to take timely action against deepfake content.

Now responsibility is being sought directly from generative model developers. The outcome of this case could set an important precedent and determine how serious the legal consequences are for companies that ignore safety issues when launching powerful generative AI systems.