Grok Generated and Distributed Nonconsensual Sexualized Images
Grok, an AI image generation tool distributed through X, reportedly generated and spread nonconsensual sexualized images of real people including minors. British lawmaker Jess Asato sued xAI alleging that Grok was designed with permissive content policies allowing such content, and sought a court order to impose technical restrictions.
What happened
Grok, an AI image generation tool distributed through X, generated nonconsensual sexualized images of real people, including minors. Users created fake sexualized images of British Labour Party member Jess Asato, including video content, after she publicly criticized Musk and Grok.
According to employees, Grok's image editing capabilities were designed to allow users to undress people in photographs. One employee estimated the bot complied with such requests approximately 90 percent of the time starting in spring 2025. Grok's internal prompts stated there were 'no restrictions on fictional adult sexual content with dark or violent themes.'
Reported impact
Asato filed a lawsuit at London's High Court alleging that Grok's design and training enabled it to generate sexualized content without consent. She is seeking a court order requiring xAI to implement technical measures to prevent Grok from creating manipulated images of her.
Users on X flagged that Grok was responding to requests to undress or edit photos of real women. One post in June attracted more than 27 million views. X's content moderation systems were ill-equipped to detect AI-generated nonconsensual intimate images.
Index analysis
Grok lagged behind OpenAI's ChatGPT in app store rankings, but the undressing controversy brought it to wider public attention. Grok vaulted to the top of app store rankings in various regions in early January 2026 following the controversy, and Musk publicly celebrated the rankings increase on X.
Safety teams at X found it difficult to identify which xAI team to contact with concerns about nonconsensual images. Key responsibilities for preventing harms rested with a small group of senior leaders overseeing product safety and model behavior, three of whom announced departures in early December 2025.
Governance controls
In mid-January 2026, xAI restricted image editing in Grok and blocked users from generating images of people in revealing clothing, following the outcry over nonconsensual sexualized images. However, the record does not specify when these restrictions were fully implemented or their effectiveness.
Asato's lawyers argued that Grok's designers made specific choices about content restrictions, including permitting adult sexual content with dark or violent themes, and that those choices should carry legal consequences.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.