Elon Musk's artificial intelligence company, xAI, faces a landmark class-action lawsuit in California alleging its Grok chatbot was trained on child sexual abuse material (CSAM) and used to generate new illegal images of a real-world survivor, Jane Doe. The suit claims xAI's data ingestion policies allowed public X posts and Grok outputs to retrain the model on abuse material. While xAI has sued its own users to shift liability, the company faces multiple class actions over its alleged failure to implement industry-standard safeguards.
Grok trained on CSAM
- ▪A proposed class-action lawsuit filed in a California federal court by a plaintiff identified as Jane Doe alleges that Elon Musk's artificial intelligence company, xAI, trained its Grok chatbot on child sexual abuse material.
- ▪The lawsuit filed by Jane Doe alleges that xAI's training policy permits public posts on X and outputs generated by Grok to be used as training data, resulting in generated child sexual abuse material being fed back into the model.
AI-generated child abuse images
- ▪Researchers at the Center for Countering Digital Hate reported that loose safeguards on Grok at the start of 2026 allowed the chatbot to generate millions of sexualized images, including thousands depicting children.
- ▪The lawsuit filed by Jane Doe alleges that Grok used real, circulating images of her childhood abuse to generate new, illegal pornographic images depicting her.
- ▪The plaintiff Jane Doe was pre-school-aged when she was abused by an adult man, and the resulting child sexual abuse material has been in circulation online for approximately 20 years.
xAI suing its users
- ▪In late August 2026, xAI filed lawsuits against its own users, Terry Wayne Harwood and Bloodworth, alleging breach of contract after they were accused of using Grok to create child sexual abuse material.
- ▪In its legal complaints against its users, xAI claimed to have assisted in the arrest of at least 244 individuals who created or distributed child sexual abuse material using Grok.
- ▪Attorneys representing victims, including Sophia Rios and Derek Potts, characterized xAI's lawsuits against its users as an attempt to deflect blame from systemic product safety failures.
CSAM hash tracking methods
- ▪The Canadian Centre for Child Protection used digital fingerprints, known as hashes, to identify AI-generated child sexual abuse material depicting the plaintiff Jane Doe on the social media platform X.
- ▪The plaintiff Jane Doe is tracked by the Federal Bureau of Investigation's Child Exploitation Notification Program, allowing her attorneys to associate the AI-generated material with a known, real-world victim.
Multiple class action lawsuits
- ▪Elon Musk's xAI is defending against at least three class-action lawsuits and multiple individual lawsuits alleging that the company failed to implement industry-standard safeguards to prevent Grok from generating non-consensual sexual deepfakes.
- ▪A class-action lawsuit filed by a plaintiff known as South Carolina Roe alleges that a perpetrator, who was her father, used Grok to generate explicit images of her as a child.
Story comments
Loading comments…