Canada's AI Strategy Prioritizes University AI Fluency

Prime Minister Mark Carney launched "AI for All," Canada's national AI strategy, in Toronto on June 4, 2026, according to the government's official announcement. The five-year strategy is built around roughly $2 billion in new federal investment and three pillars - building trust, creating opportunity, and reinforcing Canadian sovereignty - with a target of $200 billion in additional economic growth from 250,000 new AI-related jobs and raising business AI adoption from about 12% to 60% by 2034. Under the opportunity pillar, the government is establishing a National AI Literacy Initiative offering free, entry-level AI training, with a commitment to reach one million entry-level post-secondary students, train more than 3,000 educators with AI learning kits, and give every post-secondary student access to "trusted AI agents." Writing in The Conversation, Ali Shiri, a professor of information science and vice-dean at the University of Alberta, argues the strategy's literacy focus establishes broad access but does not by itself deliver "AI fluency" - the ability to critically evaluate AI systems, integrate them into disciplinary practice, and co-create with them - and reports that Canadian universities have mostly responded to AI's arrival with literacy workshops, seminars, and institutional guidelines rather than curriculum-level change.

The AI for All strategy also commits to up to 90,000 AI-related jobs and work-placement opportunities for young Canadians, a new Sovereign Technology Alliance and investment in sovereign compute, cloud and a public AI supercomputer, an AI Missions Program starting with a health-sector mission, and legislative changes covering deepfakes, AI transparency, and an expanded Canadian AI Safety Institute. It follows national consultations that drew more than 11,000 submissions and a 28-member AI Strategy Task Force. Shiri's university-fluency argument sits inside this broader push: the literacy initiative is one program within a much larger strategy, and his critique is specifically that the education piece, as designed, measures access rather than depth.

Shiri frames building fluency as requiring curriculum changes that combine conceptual foundations (how models work, data provenance, bias), applied tool training (prompting, evaluation, retrieval-augmented workflows), and project-based work with real datasets and compute access - and identifies faculty development and computational infrastructure as recurring constraints at peer institutions rather than problems that free training modules solve on their own. For anyone designing AI-literacy programs, in Canada or elsewhere, the practical takeaway is that a national access target and an institution's actual capacity to teach evaluation and responsible use are two different metrics that need to be tracked separately.

Whether federal AI for All funding flows into university curricula and faculty training specifically, rather than only headcount-based literacy modules; whether Ottawa publishes governance or safety detail for the National AI Literacy Initiative beyond its access targets; and whether the AI Missions Program's flagship health mission produces measurable results that could shape how an education-sector mission gets designed.

Shiri's fluency-versus-literacy argument is one academic's analysis, not a Canadian government position or a claim the strategy itself makes. But the pattern he describes - national workforce-training programs setting access and headcount targets that are easier to fund and report on than curriculum-depth targets, which require sustained faculty capacity most federal programs do not directly fund - is a recurring one in large-scale digital-skills initiatives generally, and is likely to resurface as other governments roll out comparable AI-literacy commitments.