Grok, the AI image-generation tool from xAI, owned by Elon Musk, violated Canada’s federal privacy law by allowing users to create and share deepfake pornographic images without consent, according to a report released Thursday by Privacy Commissioner Philippe Dufresne.
The report follows an investigation launched in January 2026, around the same time xAI implemented changes to prevent Grok from editing photos of real people into sexually explicit content.
“xAI violated Canada’s private-sector privacy law when it launched the Grok image-generation tool without deploying appropriate safeguards from the outset,” Dufresne told a news conference.
However, the commissioner lacks authority to impose financial penalties or mandate policy changes on xAI, a subsidiary of SpaceX, which is expected to launch the largest initial public offering in modern U.S. history on December 12.
xAI has pledged to actively monitor for pornographic deepfakes before receiving incident reports, rather than reacting after the fact, according to Dufresne.
The oversight report was released alongside a new digital safety bill targeting children. If passed, the law would ban children under 16 from using social media, except for companies that meet safety standards. It would also create a digital regulator to set safety standards for AI chatbots like Grok.
Global Scrutiny
xAI has faced global criticism over pornographic images on its platform. In early December 2026, British lawmaker Jess Asato filed a lawsuit against xAI related to deepfake pornographic images of herself created on the platform.
In January 2026, British media regulator Ofcom launched an investigation to determine whether the platform adequately prevented the creation of deepfake pornography.
That same month, the European Commission condemned the spread of pornographic content on X, calling it “disgusting” and “vile,” and opened a formal investigation.
In February 2026, Spain launched a probe into Grok, and in March 2026, a Dutch court ordered xAI to stop allowing the creation of nude images on its territory.
In the United States, three teenage girls filed a class-action lawsuit in March 2026, accusing the platform of enabling child sexual abuse material. The victims’ lawyers, who are remaining anonymous, argued in the complaint filed in California that xAI “had made pornography part of Grok’s DNA.”
Also in January 2026, the U.S. Senate passed a bill allowing victims of deepfake pornography to sue creators for at least $150,000 in damages.
Meanwhile, Indonesia and Malaysia have completely blocked Grok since January 2026 due to sexually explicit AI images.
