Elon Musk’s artificial intelligence company xAI faces a federal lawsuit filed on Wednesday alleging that its Grok model was trained on child sexual abuse material, including images of a survivor who was abused as a preschooler. According to court documents reported by outlets including Ars Technica and CyberScoop, the proposed class action claims the system utilized illicit content to build its image generation capabilities.
The Jane Doe Complaint and Training Allegations
The plaintiff, identified in court records as Jane Doe, was a victim of abuse in the early 2000s. Her childhood images were previously cataloged and hashed by the National Center for Missing and Exploited Children, according to the lawsuit. The Canadian Centre for Child Protection notified the plaintiff that it had identified AI-generated child sexual abuse material depicting her on xAI’s platform, causing severe psychological re-traumatization, as reported by Ars Technica. The complaint asserts that xAI incorporated material from the National Center for Missing and Exploited Children Hash List directly into its training datasets.
Platform Mechanisms and Data Recycling
The lawsuit details how xAI’s default terms treat public X posts and Grok’s own synthetic outputs as active training data. According to reporting from CyberScoop, this design means illicit material generated by the AI model flows directly back into the training pipeline. Legal filings emphasize that completely erasing the mathematical influence of a specific training example from a pre-trained neural network is technically difficult. Consequently, any illicit imagery ingested before a manual takedown continues to shape the model’s outputs even after the original files are removed from public view, as noted by Ars Technica.
Legal Demands and Statutory Claims
The proposed class action accuses xAI of violating federal child pornography statutes and Masha’s Law, which give survivors the right to sue over the production, possession, and distribution of abusive material. The plaintiff’s legal team is asking the court for monetary damages for every survivor whose childhood images were utilized in the datasets. Additionally, the lawsuit demands the complete destruction of all Grok-generated child sexual abuse material stored on xAI servers and a permanent injunction blocking Grok from generating any sexualized outputs or non-consensual intimate imagery. Representatives for X did not respond to requests for comment from reporting outlets regarding the allegations.
