Amended lawsuit alleges xAI's Grok generated over 7,000 abuse images from one childhood photo

An amended class-action lawsuit alleges xAI’s Grok chatbot was used to generate more than 7,000 sexually explicit images and videos of a Wyoming woman from a single childhood photo.

The suit, which names xAI and now adds image-generation company Stability AI as a defendant, is among the first to test whether AI developers can be held liable when their tools are used to produce child sexual abuse material, or CSAM.

According to the complaint, the plaintiff’s stepfather uploaded a photo of her taken when she was about 11 and used Grok to create the images, which were shared on social media; the stepfather later died by suicide. The suit alleges xAI reported only the original photo to the National Center for Missing and Exploited Children and did not hand over the thousands of Grok-generated images or the identifying data that law enforcement had requested.

A second new plaintiff, also identified under a pseudonym, alleges Grok was similarly used to generate abuse material from her own photos. Both new plaintiffs are named in the filing as Jane Does.

Originally filed in March and amended in July, the suit adds Stability AI as a defendant, alleging its Stable Diffusion model was trained on data containing CSAM and that safety guardrails were rolled back after user complaints. The claims sharpen a broader legal question over how much responsibility AI companies bear for downstream misuse of generative image tools.

The allegations are unproven and have not been tested in court. Both companies did not respond to requests for comment.

Whether a court lets the case proceed against the developers, rather than only the individuals who created the images, could set an early marker for AI liability in abuse cases.