A victim of childhood sexual abuse has sued Elon Musk’s artificial intelligence company, alleging that its chatbot used photographs of her abuse to generate new, illegal pornographic images depicting her.

“Using real images of the plaintiff and class members, Grok generated child pornography depicting the plaintiff and class members,” the lawsuit states. It was filed last week in the US District Court for the Northern District of California, according to the Guardian.

The plaintiff’s lawyers, who refer to her in the lawsuit as “Jane Doe” to protect her identity, accuse xAI both of generating child sexual abuse material depicting their client and of incorporating child sexual abuse material depicting her into the company’s training datasets.

The lawsuit claims that xAI ignored industry-standard safeguards, allowing the AI-generated child sexual abuse material to appear and spread on X.

Although Musk denied in January knowing that Grok had generated “nude images of minors,” his company filed a lawsuit at the end of August against two users who are facing criminal charges for allegedly doing exactly that. xAI is seeking to have the two men cover the legal costs of the lawsuits their victims have brought against the company.

This case differs from several other child sexual abuse lawsuits against xAI, as it accuses the company’s AI of using pre-existing child abuse material to generate and distribute AI-produced images. Earlier lawsuits, including one involving a group of teenagers from Tennessee, accuse Grok of using photographs of minors to create child sexual abuse material, typically by digitally removing their clothing.

The plaintiff’s lawyers said the Canadian Centre for Child Protection confirmed that hash values from the images were used to trace AI-generated child sexual abuse material depicting their client on X.

The plaintiff was of preschool age when she was raped repeatedly by an adult, according to the complaint. This series of material has circulated online for around 20 years, her lawyers said, and is known to the authorities and to child protection organisations.

“The difference here, compared with many of the criminal cases you’ll see involving AI-generated material, is that you can’t prove the child in the material is real. But here, because you can trace it back to the series, we know this is an identifiable victim who is still alive and was a real child in the photograph,” said Margaret Mabie, the plaintiff’s lawyer.

It is worth noting that xAI is facing numerous lawsuits over the creation and distribution of child sexual abuse material. Many of these stem from a period in early 2026, when lax safety mechanisms on Grok allowed users to instruct it to remove clothing from photos and generate sexual images on command via X. Musk said in January that Grok’s NSFW function was intended to allow “upper-body nudity of fictional adult people.” The bot generated around 3 million sexualised images over an 11-day period, including thousands depicting children, according to researchers at the Center for Countering Digital Hate.

The use of AI to generate child sexual abuse material presents a new range of threats to victims’ safety and mental health, the lawyers said, exposing them to unpredictable harm and retraumatising them with every new image created.

“The problem here is that, at least for our client, the number of images depicting her over the past 20 years was limited,” Mabie said. “The fear now is that new criminal acts, new offensive behaviour, new ideas about what could happen to them, can now be generated using AI. They can cause new abuse,” she added.