This House of Commons Science and Research committee meeting studied the development of a responsible artificial intelligence research ecosystem in Canada. Witnesses included Professor Mehmet Murat Kristal, Dr. Taylor Owen, Dr. Steven Murphy, Dr. Peter Lewis, Jim Hinton, Anne Nguyen, and Dr. Tijs Creutzberg.
Professor Kristal argued that Canada's weakness is not in AI research or talent but in translating that excellence into large-scale deployment and productivity gains. He recommended building AI literacy among executives and regulators, creating shared national data infrastructure, and establishing clear deployable standards for trustworthy AI, warning that treating AI as only an innovation issue rather than a national infrastructure issue would be a major policy mistake.
Dr. Owen stated that a responsible AI research ecosystem requires a responsible AI governance ecosystem, and that public confidence is currently lacking. He recommended closing the governance gap through action on citizen safety, information integrity, and democratic legitimacy, arguing these measures could be implemented through an amended online harms act and an amended consumer privacy protection act, and that sustained investment in social science and interdisciplinary research is needed.
Dr. Murphy argued that Canada's opportunity lies in developing trustworthy, human-centred AI applications in sectors that contribute most to Canadian GDP, such as energy, mining, and advanced manufacturing, rather than trying to compete with U.S. hyperscalers on large language models. He commended government efforts to recruit international researchers but stressed the need to hire across the full gamut of AI-related disciplines.
Dr. Lewis argued that AI research and innovation must be deeply interdisciplinary and that Canada should diversify its efforts beyond machine-learning institutes. He recommended a values-first approach to AI leadership, focusing on dignity, inclusivity, equity, and truth rather than speed, and suggested Canada should invest in the seeds of what comes next after the large language model boom, disagreeing with the notion that competing to build the biggest models is the right path.
Jim Hinton argued that Canada has no actual AI strategy and is doing a poor job of owning AI, with its share of AI patent ownership dropping since 2017. He recommended stopping all funding of talent and research that gets owned by foreign companies, building truly sovereign compute infrastructure out of reach of foreign control, and spurring an IP economy, stating that Canada cannot commercialize or govern what it does not own and control.
Anne Nguyen recommended that the government lead by example and become a customer of publicly funded innovation, create demand by investing upstream in recurring matters of public interest, and make AI literacy a national jurisdiction. She also recommended structuring an industrialization continuum from basic research to commercialization and turning knowledge into a public good through open innovation infrastructure like the collective intelligence space Brigade IA.
Dr. Creutzberg reported that indicators continue to point to long-standing challenges in capturing the benefits of homegrown discoveries, including scaling start-ups' access to capital and low business adoption of AI. He noted that AI blurs disciplinary boundaries, challenging how research is administered and funded, and flagged the need for researchers to embrace a modern mindset around dual-use risks and security responsibilities, while cautioning that AI tools can amplify bias and miss important contextual knowledge.
AI-generated summary — may contain errors; verify against the official evidence.