The House of Commons Access to Information, Privacy and Ethics Committee heard from four witnesses on February 2, 2026, as part of its study on the challenges posed by artificial intelligence and its regulation. The witnesses were David Krueger, an assistant professor at Université de Montréal; Anthony Aguirre and Max Tegmark of the Future of Life Institute; and Privacy Commissioner of Canada Philippe Dufresne.
David Krueger testified that superintelligent AI could cause human extinction within roughly five years, that researchers do not know how to control or safely test such systems, and that AI companies are effectively "rolling the dice" by racing to build them. He recommended an immediate international pause on developing more powerful AI, arguing that the concentrated supply chain for advanced computer chips makes such a pause technically feasible, and said this should be Canada’s top foreign policy priority.
Anthony Aguirre warned that AI companies are aiming to build recursively self-improving superintelligence that would operate at inhuman speed and be impossible to predict or control, with many builders estimating a 10% to 20% risk of human extinction. He recommended that Canada support "tool AI" designed to empower rather than replace humans, require that all AI systems remain under meaningful human control, and prohibit the development of superintelligence until it can be shown safe and there is broad public buy-in.
Max Tegmark argued that AI is currently less regulated than sandwiches in Canada, allowing products like chatbots linked to child suicide to be sold without predeployment testing. He recommended that Canada treat AI companies like pharmaceutical companies, requiring clinical trials to prove benefits outweigh harms before any product is released, starting with child safety, and said this would create a golden age of safe, innovative AI in Canada rather than driving companies away.
Philippe Dufresne stated that privacy must be at the heart of AI regulation, noting that 83% of Canadians are concerned about privacy when using AI tools and that his office has investigated TikTok, OpenAI, and X’s Grok chatbot. He recommended that Parliament modernize Canada’s privacy laws by granting his office order-making and fine-imposing powers, recognizing privacy as a fundamental right, requiring privacy-by-design and impact assessments for high-impact data processing, and creating a children’s privacy code, adding that without these tools he cannot compel companies like Aylo (Pornhub) to stop non-consensual image sharing.
In response to questions, Krueger and Aguirre both said that AI companies acknowledge the extinction risk but feel trapped in a race they cannot stop alone, and that only government intervention can defuse it. Krueger disagreed with any suggestion that the risk is overstated for corporate hype, calling that claim demonstrably false. Dufresne confirmed he was not consulted on the government’s deal to import Chinese electric vehicles, said he has been consulted on updated PIPEDA legislation and is optimistic about gaining order-making powers, and noted that Bill C-8’s cybersecurity powers lack a necessity-and-proportionality standard and judicial authorization, which he has recommended be added.
The committee briefly discussed procedural matters, including seeking House concurrence to study the Conflict of Interest Act and the Lobbying Act, and noted that the AI minister will appear for one hour on Thursday followed by officials, despite the committee’s request for two hours.
AI-generated summary — may contain errors; verify against the official evidence.