A lawsuit filed in a California federal court has accused Elon Musk’s artificial intelligence company xAI of using decades-old images of a child sexual abuse victim to generate new abusive material through its Grok chatbot.

The complaint, filed Wednesday in the Northern District of California by an anonymous plaintiff identified as Jane Doe 1, alleges that xAI designed Grok to transform real photographs of identifiable people into sexualized content and publish the resulting images directly on the X social media platform, formerly known as Twitter.

According to the lawsuit, the plaintiff was identified as a victim by the US National Center for Missing and Exploited Children in the early 2000s. Images documenting her abuse have circulated online for nearly two decades.

The complaint alleges that Grok used that existing material to create new imagery depicting the plaintiff. The AI-generated material was subsequently identified by the Canadian Centre for Child Protection, according to the filing.

The lawsuit seeks damages under Masha’s Law, a US statute that allows victims of certain federal child pornography offenses to recover at least $150,000 per violation. It also asks the court to order the destruction of any illegal material held by the company.

The complaint links the allegations to a surge in Grok-generated sexual imagery that began in late December, after Musk asked the chatbot to depict him wearing a bikini.

Citing research by the Center for Countering Digital Hate, the lawsuit claims Grok generated more than three million sexualized images during the 11 days leading up to January 8, including more than 23,000 images that appeared to depict children.

The lawsuit alleges that xAI initially responded to concerns by restricting image-generation capabilities to paying subscribers rather than disabling the feature altogether.

The complaint also contrasts xAI’s approach with those of OpenAI, Anthropic and Meta, which generally restrict sexual image-generation requests. It further points to xAI’s promotion of a “spicy mode” for its Grok video tool.

The case is at least the second class-action lawsuit accusing xAI of allowing Grok to generate abusive material.

A separate lawsuit filed this year by three teenagers in Tennessee has since been expanded to include additional plaintiffs and to name image-generation company Stability AI as a defendant.

SpaceX, which is identified as xAI’s parent company in the complaint, did not immediately respond to a request for comment.