Key facts
- xAI is accused of training its Grok models on child sex abuse material (CSAM).
- The lawsuit claims that AI-generated CSAM depicting the plaintiff, Jane Doe, was used in training data.
- Doe alleges that xAI's terms do not exclude CSAM, non-consensual intimate imagery (NCII), or NSFW material from training data.
- The lawsuit accuses xAI of violating federal child pornography laws and Masha's Law.
- The plaintiff seeks monetary damages and an order for xAI to destroy CSAM and prevent its generation.
xAI, the artificial intelligence company founded by Elon Musk, is facing a lawsuit alleging that its Grok models were trained on child sex abuse material (CSAM). The complaint, filed by a plaintiff identified as Jane Doe, claims that her images, including AI-generated CSAM depicting her, were part of the datasets used by xAI.
Doe alleges that her images, which were created when she was a preschool-age victim of repeated rape in the early 2000s, have been hashed by organizations like the National Center for Missing and Exploited Children (NCMEC) and the Canadian Centre for Child Protection (CCCP). She was notified by the CCCP that AI-generated CSAM depicting her had been identified on xAI's platform. The lawsuit states that messages were found on online forums between offenders discussing the creation of AI-generated CSAM of "similarly situated known, legacy, victims of CSAM."
The complaint further alleges that xAI not only made it easier to create violative images but also stores Grok's outputs and uses them to further train the model. This means Grok may have been trained on both the original images and the subsequent AI-generated ones. While xAI reportedly filters out violent content from its training data, the lawsuit notes that its terms of service do not specify whether CSAM, non-consensual intimate imagery (NCII), or NSFW material are excluded categories. The difficulty in removing a training example's influence from an already-trained model suggests that any CSAM ingested before takedown likely continued to shape Grok's outputs.
The proposed class action seeks to halt Grok's harmful outputs and represents victims whose childhood images may have been used. The lawsuit accuses X (formerly Twitter) and xAI of violating federal child pornography laws and Masha's Law, which allow CSAM survivors to sue for production, possession, and distribution. Lawyers for Doe stated that xAI allegedly committed all three offenses. If successful, xAI could owe damages to victims whose real photos were used to generate CSAM via Grok. Doe is also requesting the court order xAI to destroy any stored CSAM and prevent Grok from generating such content in the future, which her complaint suggests might require blocking all sexualized outputs, including NCII and NSFW material.