Science and Research

This House of Commons Science and Research committee meeting studied artificial intelligence, hearing from professors Pina D'Agostino and Arvind Gupta, University of British Columbia Vice-President Gail Murphy, Canadian Nuclear Association President George Christidis, Electricity Canada President Francis Bradley, and Ontario Power Generation Vice-President David Donovan.

Professor D'Agostino argued that AI sovereignty must address ownership, access, and control of intellectual property and data, not just capacity. She warned that Canadian-generated IP often flows to foreign multinationals, and recommended a three-part strategy of creating and protecting IP in Canada, keeping control domestically, and scaling through global companies. She noted that demand for accessible IP services outpaces supply, and stressed that AI systems must not extractively use creators' copyright-protected materials without permission and fair compensation.

Professor Gupta reported that consultations for the federal AI strategy task force emphasized the need for mission-driven research, talent development, and a national focus on strategic sectors like defence, resource extraction, health care, and microelectronics. He cautioned against mimicking other countries' strategies and recommended sovereign compute for advanced research while using shared compute for corporate needs. He also highlighted the importance of building public trust through ethical AI policies and ensuring Canadians see tangible benefits from AI.

Professor Murphy identified four key areas to support AI research translation: enhancing AI institutes as hubs for talent and mission-driven challenges, focusing on areas of global strength like AI for health and robotics, establishing precompetitive industrial research centres for critical parts of the AI supply chain, and building large-scale sovereign compute infrastructure. She noted that Canadian industry has one of the lowest rates of AI adoption globally and that barriers include difficulty accessing diverse data and insufficient funding for collaborative projects.

Mr. Christidis emphasized that nuclear energy provides the predictable, carbon-free baseload power that AI data centres require, and that Canada's nuclear industry is a national asset with world-class uranium resources and reactor technologies. He argued that timely regulatory processes, federal-provincial collaboration, and sustained investment in nuclear innovation are essential to meet rising electricity demand from AI, and noted that the AI imperative reinforces the urgency of nuclear expansion.

Mr. Bradley stated that electricity demand from data centres is projected to double by 2050, requiring more generation, transmission, and distribution infrastructure in 25 years than in the last century. He recommended eliminating barriers to building critical electricity infrastructure, facilitating investments, including the electricity sector in AI policy discussions from the start, and protecting critical infrastructure from AI-related cyber-threats. He noted that no electricity sector representative is on the federal AI task force, which he called a significant gap.

Mr. Donovan highlighted that Ontario's grid, over 80% non-emitting, positions Canada to lead in AI, but that data centre loads arrive in massive increments of up to 1,000 megawatts each. He noted that Ontario's Independent Electricity System Operator projects a 65% increase in electricity need by 2050, with data centres requiring about 1,600 megawatts by 2040. He called for proactive planning and regulatory processes now, given long lead times for nuclear and hydro projects, and welcomed federal support through investment tax credits and low-cost financing.

The committee also debated a motion from Mr. Blanchette-Joncas to invite representatives of the three federal granting councils to appear regarding data governance and funding equity, and agreed to a wording change from Mr. Baldinelli to request a government response under Standing Order 109.