Canada’s Privacy Commissioner, Philippe Dufresne, released findings on June 11 concluding that Grok, the AI chatbot developed by Elon Musk’s xAI, violated Canadian federal privacy law by generating non-consensual intimate deepfake images. The investigation revealed that over 1.8 million sexualized deepfake images were created globally through Grok in just 10 days earlier this year. The Commissioner noted that while X (formerly Twitter) and xAI have implemented some safety measures, the issue is far from resolved, as users can still produce and share unauthorized pornographic content using Grok.
The investigation, launched in January 2026, ran concurrently with similar probes in the United Kingdom and California. Canada’s privacy commissioner focused on whether the companies obtained “valid consent” when collecting, using, and disclosing personal data to generate deepfakes. Dufresne stated that Grok’s image-generation tool was launched without sufficient safeguards and without adequate consideration of potential privacy harms. Although X and xAI have committed to fixes, the Commissioner said “the issue has not been satisfactorily resolved.”
