Suit says xAI hindered police after Grok abuse-image report
An amended class action says Grok was used to create abuse images of minors and that xAI failed to provide key data to investigators.
By Maya Lindqvist · Senior Technology Correspondent
3 min read
An amended proposed class action accuses X and xAI of allowing Grok to generate AI child sexual abuse material and then withholding information police needed to identify users. The filing argues the alleged failure turned a product-safety problem into an obstacle for law enforcement investigating crimes against children.
The complaint, expanded Tuesday, says one girl’s stepfather used a single photo of her taken when she was 11 to create about 7,000 sexually explicit AI images and videos through Grok. According to the lawsuit, xAI’s system reported only after the man entered a prompt involving “gang rape,” sending a CyberTip to the National Center for Missing and Exploited Children.
The complaint says xAI did not provide investigators with basic user-identifying information, including an IP address, despite mandatory reporting rules for flagged child sexual abuse material. Lawyers for the plaintiffs said in a press release that xAI’s report to NCMEC included the original non-explicit photo, did not include the AI-generated images, and omitted location information that investigators repeatedly sought.
Police later arrested the man after obtaining a warrant to seize his devices, according to the amended complaint. A forensic review allegedly found roughly 7,000 AI-generated images and videos of the girl, identified in the suit as Jane Doe 4. The complaint says the man also traded the material online for child sexual abuse material made by others.
Two days after his release on bail, the man died by suicide, according to the complaint. The lawsuit says Jane Doe 4 has since experienced anxiety, depression and suicidal thoughts.
Lawyers allege a broader reporting pattern
The plaintiffs’ law firms, Lieff Cabraser Heimann & Bernstein and Baehr-Jones Law, said NCMEC found in early 2026 that 90% of xAI’s CyberTipline reports were not usable by police because xAI did not include information needed to track suspects. Jane Doe 4 said in the firms’ press release that X had enough information to help police but “remained silent.”
The amended complaint also adds another minor, Jane Doe 5, who allegedly was targeted by an adult family friend. The filing says police notified her after the suspect was arrested and charged over possession and distribution of child sexual abuse material, though those charges were limited to non-AI material.
Elon Musk, xAI’s founder, has denied that Grok has been used to create child sexual abuse images. Ars Technica reported that X and xAI did not respond to requests for comment.
Stability AI added as defendant
The amended complaint also names Stability AI as a defendant. The suit alleges that Stability AI’s open-weight models underpin third-party “nudify” apps and that some perpetrators targeting plaintiffs may have used Grok, Stability AI-linked tools, or both.
The complaint cites a June research report that found the Stable Diffusion model family accounted for 42.7% of image-based online nudification. Stability AI did not respond to Ars Technica’s request for comment, according to the report.
The lawsuit now seeks classes covering people in the United States whose images as minors were altered into sexually explicit content through Grok or through apps built on Stability AI models. The plaintiffs’ lawyers estimate that thousands of minors may qualify.
NCMEC warned in March of a sharp increase in reports involving generative AI and child sexual exploitation. The group said more than 1.5 million CyberTipline reports in 2025 indicated a connection to generative AI, while more than 133,000 lacked enough information to determine how the technology was used.
This story draws on original reporting from Ars Technica.