Prime Minister Mark Carney’s Liberal government has tabled its first official budget, entitled Canada Strong. It frames itself as a road map of investments being made to strengthen national sovereignty via economic productivity and national defence. Central to those efforts is artificial intelligence.AI-heavy technologies have been identified by eight federal agencies in the 2025 budget as a way to reduce operational expenditures while fuelling productivity.Many of the investments in the budget are aimed at developing the defence industry through the creation and commercialization of what’s known as dual-use technology — goods, software and technology that can be used for both civilian and military applications — which can also include AI. But is Canada Strong actually weak on AI?Given the current legislative landscape and the new budget, we argue that Canada Strong’s AI plan downplays regulation and guardrail development, since funding is geared chiefly toward adoption. It overlooks the risks, impacts and potential weaknesses that come with an over-reliance on these technologies.Past budgetsIndirectly, the Canadian government has consistently supported AI research through the Federal Granting Agency, the Canada Foundation for Innovation and the Canadian Institute for Advanced Research. Between 2006 and 2015, Prime Minister Stephen Harper’s government invested more than $13 billion in science, technology and innovation during its mandate.Justin Trudeau’s government changed how AI was marketed to Canadians and how it was funded. The 2017 budget, entitled Building a Strong Middle Class, made the first explicit references to AI in a federal budget, describing it as representing a transformative force for the Canadian economy. The government emphasized “Canada’s advantage in artificial intelligence,” which it said could translate into “a more innovative economy, stronger economic growth, and an improved quality of life for Canadians.”Bill Morneau, the finance minister at the time, proposed funding AI superclusters and allocating $125 million to establish the first Pan-Canadian Artificial Intelligence Strategy. This commitment to AI was reaffirmed in the 2021 budget, when the technology was presented as “one of the most significant technological transformations of our time.” The federal government’s investments in the sector were portrayed as essential to ensure the economy benefited, and that Canada’s position of strength enabled the “integration of Canadian values into global platforms.” The government renewed the Pan-Canadian AI Strategy with another $368 million. An additional $2.4 billion was committed in the 2024 budget, which emphasized the “safe and responsible use” of AI, notably through the creation of new standards and the establishment of a Canadian AI Safety Institute Sovereignty focusThe 2025 budget marks another substantial shift in Canada’s approach to AI. This third phase of funding focuses on adoption, productivity, sovereignty and the fundamental principle of dual use, both civilian and military.But we don’t believe it fosters research and projects addressing the key issues tied to AI, and instead amplifies promotional language.We believe the large-scale adoption of AI across federal departments and agencies (like the Canada Revenue Agency, Employment and Social Development Canada, Fisheries and Oceans Canada, Public Services and Procurement Canada, Statistics Canada, Natural Resources Canada and Canadian Heritage) will actually reduce the capacity to pursue regulatory development, guardrail design, ethical deliberation and meaningful civil-society input because its widespread integration will permeate the entire bureaucracy.AI presented as an economic driver through cost reduction and dual-use applications has become the new promotional narrative for the government. Read more: What are Canada's governing Liberals going to do about AI? The AI weakness in Canada StrongWhat vulnerabilities arise when AI is aggressively deployed within the public service? Since the abandonment of the Artificial Intelligence and Data Act in 2025, Canada’s approach to AI governance has relied more on norms and standards than on the rule of law. This environment could risk overturning a perceived AI advantage into one of weakness. This is especially true given an over-reliance by the government on foreign software (such as Microsoft CoPilot) and hardware (NVIDIA chips needed for super computers), a lack of comprehensive understanding of the technologies already in use by different agencies and no guidelines on lethal autonomous weapons — weapons systems that can independently search for, identify and attack targets without direct human intervention. Read more: How Russian and Iranian drone strikes further dehumanize warfare Promoting rapid regulatory design and AI adoption within a budget focused on stimulating dual-use research, development, commercialization and implementation risks overlooking many of AI’s pitfalls, including:Reduced administrative transparencyEmbedded discriminationModels unsuited to administrative or military standardsMass surveillance, unethical use and unintended collateral damage.Promotion AI as an economic boon — through public administration automation and military dual use — within an unregulated environment, and without dedicated funding for oversight, risks disrupting key sectors and services that sustain Canadian democracy, the very foundation of “Canada Strong.”Nicolas Chartier-Edwards receives funding from the Social Sciences and Humanities Research Council.François-Olivier Picard does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.