Deepfake Images Generated by Grok Spark Privacy Investigation in Canada and Other Countries

Canada’s Privacy Commissioner, Philippe Dufresne, has revealed that sexually explicit deepfake images generated through Grok, the AI chatbot developed by Elon Musk’s artificial intelligence company xAI, violated the country’s federal privacy laws. According to the findings of the investigation, the chatbot was, at one point, generating more than 6,000 sexually explicit deepfake images every hour, raising serious concerns about privacy and online safety. According to the investigation report released by Commissioner Dufresne, xAI and X Corp. failed to implement adequate safety measures before launching Grok’s image-generation tool. As a result, users were able to create sexually explicit deepfake images of individuals without their consent and easily distribute them on the social media platform X (formerly Twitter). Researchers involved in the investigation stated that the technology was misused to create millions of deepfake images, with women and children being the primary victims. The investigation was launched in January this year to determine whether the companies had complied with Canada's federal privacy laws while operating the AI tool. Commissioner Dufresne emphasized that victims of deepfake technology often suffer severe psychological and social consequences that can persist for a long time. He stressed that Canadians, particularly children, deserve a safe online environment where their privacy is protected. Following the launch of the investigation, xAI and X Corp. introduced several improvements to their systems aimed at preventing the misuse of real individuals’ images and enabling the rapid removal of harmful content. The companies have also agreed to submit quarterly progress reports and undergo independent third-party audits. The issue has drawn international attention, with similar investigations into Grok underway in the United Kingdom, the European Union, and the U.S. state of California. Meanwhile, Canada’s Liberal government has introduced a legislative proposal to classify the creation of non-consensual sexually explicit deepfake images as a criminal offence. Concluding his remarks, Commissioner Dufresne said that Canada’s federal privacy laws must be modernized and strengthened to keep pace with evolving technology and to ensure strict action can be taken against companies that fail to comply with privacy regulations.