Grok Chatbot Found to Have Breached Canada's Privacy Law

Canadian Privacy Commissioner Philippe Dufresne has found that xAI's chatbot Grok violated federal privacy rules through the launch of an image-generating tool that allowed sexualized deepfakes of real people to be created and shared without their consent.

The investigation began in January. xAI subsequently modified the service to restrict the editing of photos of real people into sexually explicit content. According to Dufresne, the company had not put sufficient safeguards in place when it launched the tool.

The commissioner, who does not have the power to impose fines or order changes, praised the steps taken by xAI, saying the company had committed to actively monitoring the occurrence of sexualized deepfakes rather than waiting for users to report problems.

The Elon Musk-owned company did not immediately respond to a request for comment.

(reuters, bak)