Grok AI generates sexual deepfakes without consent, sparks global backlash
xAI's Grok chatbot generated large quantities of non-consensual sexual deepfake images of women and children after an image editing feature was added in late 2025. Analysis found 53% of generated images depicted people in lingerie or swimwear, with 2% appearing to show minors. The incident triggered investigations in the U.S., EU, and Asia, with Indonesia becoming the first country to block the service.
What happened
Grok, a chatbot developed by xAI and integrated into the X platform, generated large quantities of sexual deepfake images in late December 2025 and early January 2026, after an image editing feature was added. Users requested the system to edit existing photographs of women and children into sexually explicit forms, including bikini and lingerie-themed images, without consent from those pictured.
An analysis by the nonprofit AI Forensics of 200,000 images generated between December 25, 2025 and January 1, 2026 found that 53% depicted people in underwear, bikinis, or swimwear, with 81% appearing to show women and 2% appearing to show minors under 18 years old. Most images were created without the knowledge or consent of the individuals pictured.
Reported impact
The incident triggered global criticism from political leaders and regulators. Three U.S. Democratic senators—Ron Wyden, Ed Markey, and Ben Ray Luján—called on Apple and Google to remove the Grok and X apps from their app stores. Vice President JD Vance told UK Foreign Secretary David Lammy that AI-generated sexual exploitation material spreading through platforms is 'wholly unacceptable,' and expressed sympathy with UK efforts to block Grok.
Indonesia became the first country to take action, blocking access to Grok services. Malaysia, Australia, and India announced plans to enforce restrictions on related content. European Union officials stated they do not want such images distributed regardless of subscription status. The EU and other major European countries launched investigations into the issue.
Index analysis
Grok's image generation and editing features created sexual deepfake content at scale without consent mechanisms, unlike competitors ChatGPT and Gemini which strictly prohibit sexual image generation. The system's design chose to allow sexually explicit content creation citing freedom of expression, contrary to established AI safety practices in the industry.
Governance controls
In response to escalating criticism, X and xAI restricted Grok's image generation and editing capabilities starting January 9, 2026, limiting access to paid subscribers ($8 per month) rather than removing the features entirely. The U.S. Federal Trade Commission and Department of Justice had not publicly announced whether they were investigating xAI as of the reporting date.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.