Class Action Lawsuit Accuses xAI of Training Grok AI on CSAM

Elon Musk’s artificial intelligence venture, xAI, is facing a major federal class-action lawsuit alleging that the company trained its Grok models on both real and AI-generated child sexual abuse material (CSAM).

The lawsuit marks the first legal complaint directly accusing xAI of utilizing illegal imagery of children to build the generative capabilities of its Grok model.

Accusations of CSAM in Training Pipelines

The complaint was filed by a plaintiff designated as Jane Doe, a survivor of childhood sexual abuse whose physical images have been tracked for over two decades using digital hash lists maintained by groups like the National Center for Missing and Exploited Children (NCMEC) and the Canadian Centre for Child Protection (CCCP).

According to the filing, Doe was alerted by the CCCP after the organization discovered AI-generated CSAM depicting her circulating on xAI’s platform. Online forums revealed that offenders were actively using Grok to generate synthetic abuse imagery based on known, real-world victims.

Attorneys for the plaintiff argue that xAI built Grok's image and video generation capabilities using datasets that included NCMEC-flagged CSAM. Furthermore, the suit highlights xAI’s data ingestion policies, which default to using public posts on X alongside Grok’s own generated outputs as ongoing training data.

While xAI maintains filters to exclude violent content from training pipelines, the complaint notes that its public terms do not explicitly exclude CSAM, non-consensual intimate imagery (NCII), or explicit adult content. Because removing individual data points from pre-trained AI models is technically complex, the lawsuit asserts that ingested CSAM continues to shape Grok’s outputs even after specific posts are taken down.

Legal Remedies and Demands

The lawsuit claims xAI violated federal child pornography laws and Masha's Law, which grants survivors civil recourse against the legal production, possession, and distribution of CSAM.

Attorneys Margaret E. Mabie and Sarah London stated that xAI should be held legally liable for enabling the creation and recirculation of abuse material, which subjects victims to ongoing re-traumatization.

If certified, the class action would represent all survivors whose childhood images were used to generate Grok content. The suit demands statutory damages for affected individuals and asks the court to order xAI to permanently delete all Grok-generated CSAM stored on its servers. Additionally, the plaintiff seeks an injunction forcing xAI to implement strict controls preventing Grok from generating any sexualized content.

Neither X nor xAI responded to requests for comment on the suit.