Canada’s privacy commissioner is broadening the inquiry into Elon Musk’s X Corp. after reports emerged that its AI chatbot Grok is being utilized to generate and distribute explicit photos of individuals without their authorization. Commissioner Philippe Dufresne emphasized the seriousness of using personal data to produce deepfakes, including intimate imagery, without consent, stating that it infringes on individuals’ right to privacy. Recognizing the potential harm, Dufresne has decided to extend the investigation to encompass not only X Corp., the operator of X (formerly Twitter), but also xAI, Musk’s AI company behind Grok.
The proliferation of sexual deepfakes on Musk’s platform X, produced by the Grok chatbot, has sparked global outrage in recent weeks, primarily targeting women and, in some instances, minors. While the technology for creating sexualized deepfakes is not new, X made it easily accessible by allowing users to request Grok to edit images directly on the platform.
In response to the controversy, X updated Grok to prevent the editing of images of real individuals in revealing attire like bikinis. These restrictions apply to all users, including paid subscribers, who can still edit images through their Grok accounts on the X platform, with measures in place to hold accountable those attempting to misuse the service.
Artificial Intelligence Minister Evan Solomon recently stated that Canada is not contemplating a ban on X, a comment that garnered praise from Musk. However, Malaysia and Indonesia have announced plans to block access to Grok, with a potential ban also looming in the U.K.
Dufresne’s office confirmed the expansion of the investigation to examine whether X Corp. and xAI obtained proper consent from individuals whose personal information was used to create deepfakes, including explicit content. The privacy commissioner’s office is considering X’s commitment to addressing the issue in its ongoing investigation.
The initial probe, launched in February 2025, focused on assessing X’s compliance with federal privacy laws regarding the collection, use, and disclosure of Canadians’ personal information for training AI models, under Canada’s Privacy Act.