xAI sues user for using Grok to generate nonconsensual sexualised deepfakes

xAI has initiated legal proceedings in a Texas court against Terry Wayne Harwood, a 67-year-old man from South Carolina, for the alleged misuse of its Grok artificial intelligence platform. The lawsuit centres on claims that Harwood systematically bypassed safety filters between December 2025 and February 2026 to generate nonconsensual sexualised images of adults and minors. This action follows broader regulatory scrutiny of Grok’s safety protocols, with investigations launched by authorities in California, the UK, the European Union, and Ireland after reports emerged in early January that the AI was facilitating the transformation of real individuals into sexualised content.

While xAI implemented safeguards in response to these reports, the company alleges Harwood continued to upload and edit images, including specific instances involving minors. The complaint details that Harwood uploaded non-sexual photos to two accounts and repeatedly submitted edited prompts to circumvent safeguards after Grok refused requests on numerous occasions. One specific example cited involves Harwood uploading a photo of a fully dressed girl, estimated to be 10 to 11 years old, and asking Grok to remove her clothing and make her perform a "Playboy model impression."

Harwood was arrested in March on separate state charges related to the sexual exploitation of minors. The South Carolina Attorney General’s office announced his arrest on March 9 as part of the Internet Crimes Against Children Task Force activities. Harwood faces three counts of second-degree sexual exploitation of a minor and five counts of sexual exploitation of a minor in the third degree. The lawsuit notes that Harwood did not just possess child sexual abuse materials but also distributed them through his interactions with the AI.

xAI is seeking unspecified monetary damages and reimbursement for legal defence costs. The company is asking the court to order the defendant to pay for the expenses it incurs to defend itself against potential legal action from victims. This marks one of the first lawsuits brought by an AI company against a user for misuse of its platform, sending a clear message regarding the consequences of violating terms of service.

Reports emerged in early January that Grok was allowing users to transform photos of real women and children into sexualised images. California authorities began looking into the AI in mid-January, while UK regulator Ofcom, the European Commission, and Ireland's Data Protection Commission opened separate probes. Despite these measures, users were still able to use Grok to undress men, and Harwood continued uploading and editing images.