In a plot twist that surprises absolutely no one who has been paying attention to the AI industry's 'gather all data, ask questions later' approach, Elon Musk's xAI has been hit with a lawsuit alleging that its Grok models were trained on child sexual abuse materials (CSAM). The complaint, filed on Wednesday by a plaintiff identified only as Jane Doe, claims that her childhood abuse images - captured in the early 2000s when she was preschool-age - were used to train Grok's image and video generation capabilities.

Doe's images have long been hashed by groups like the National Center for Missing and Exploited Children (NCMEC) and the Canadian Centre for Child Protection (CCCP) to help identify and remove such content. She receives alerts from the US Department of Justice Victim Notification System whenever she might be a victim in a new investigation. So you can imagine her shock when CCCP notified her that AI-generated CSAM depicting her had been found on xAI. The complaint alleges that online forums featured 'offenders chatting about creating AI generated CSAM of Plaintiff and other similarly situated known, legacy, victims of CSAM.'

Doe now fears that xAI has not only made it easier to create more images of the most distressing time in her life, but also that the company has stored and used those AI-generated outputs to further train Grok. The complaint suggests that Grok may have been trained on both the original images that have haunted her for over 20 years and the recent AI-generated ones.

This is the first case to accuse xAI of training on CSAM, but the complaint is a bit light on details for that specific claim. It notes that Doe's images were part of a CSAM Hash List maintained by NCMEC, and 'that same material' allegedly 'was part of the dataset xAI used to build Grok's image and video generating capabilities.' The complaint also points out that while xAI filters out violent content from training data, its terms don't explicitly exclude CSAM, non-consensual intimate imagery (NCII), or NSFW material. And since removing a training example's influence from an already-trained model is technically difficult, any CSAM ingested before takedown likely continued to shape Grok's outputs.

Doe's proposed class action seeks to represent every victim whose childhood images have been used to generate Grok CSAM. It accuses X of violating federal child pornography laws and Masha's Law, both of which give CSAM survivors a right to sue. Doe's lawyer, Margaret E. Mabie, suggested that 'xAI did all three' - production, possession, and distribution.

If Doe wins, xAI could owe damages to every victim who can prove Grok generated CSAM based on their real photos. Doe also wants the court to order xAI to destroy all Grok-generated CSAM and block Grok from generating any sexualized outputs, including the NSFW 'bikini pics' that Musk has promoted. Sarah London, another of Doe's lawyers, said in a press release that Doe 'has lived for nearly two decades knowing that images of the worst thing that ever happened to her are circulating among predators online, and that they can resurface at any moment. xAI must be held responsible for knowingly training its models on images of the horrific abuse she suffered, and on the abuse images of every other survivor in this class.'

X did not respond to Ars' request to comment - because why comment when you can just let the legal system sort it out?