Artificial intelligence no longer sits on the horizon of higher education. It is on the agenda of every senate meeting, budget review and curriculum committee. Ottawa is pouring money into AI research at universities and has launched a literacy strategy that promises to reach one million students entering post-secondary study. Behind the funding announcements sits an unsettled question: what does it mean for a university to be “ready” for AI?
Decades of scholarship on technology in higher education suggest that buying or licensing a tool is the easy part. Institutions change when their culture, staff expertise, decision-making structures, infrastructure and leadership change with the technology. Software alone rarely moves a university anywhere.
To see how Canadian institutions are handling this, a research team went looking. Guy Levesque of the Hunter Hub for Entrepreneurial Thinking at the University of Calgary and the author of the original study examined 15 research-intensive universities across Canada between March and June 2026. The work has not yet been peer-reviewed, and the authors say plainly that it is a provisional snapshot rather than a full census. Its purpose is comparison and provocation, not ranking.
What the scan counted
The team combed the public websites of the 15 institutions and ran keyword searches across them. The haul was 953 publicly documented AI initiatives, spanning classroom teaching, student AI literacy, new programs and credentials, entrepreneurship, partnerships with outside organizations, computing infrastructure, events, funding, student clubs and governance.
Sorting these into patterns, the researchers saw several distinct institutional “personalities.” Some universities lean into research and external partnerships. Others have built strong pipelines that carry AI ideas toward startups. Some put their energy into coordination and oversight, creating central bodies to steer the effort. Others let AI work bloom across faculties, labs and departments with little central direction. A further group has concentrated on teaching, professional development and AI literacy.
The lesson is that there is no single AI strategy on offer in Canada. Each institution is making choices about what it wants AI to do for its students, researchers and communities, and those choices reflect different readings of what a university is for.
The thin slice devoted to teachers
One number from the scan stands out. Of the 953 initiatives, only 91, about nine per cent, were aimed explicitly at instructors.
That figure matters because instructors sit at the point where AI policy turns into daily practice. They decide when a chatbot helps a student learn and when it short-circuits the learning. They have to rethink assessments in a world where essays and problem sets can be generated on demand. They have to define “responsible use” in their own course, and prepare students to know what they are handing off when they delegate a task to a machine.
None of this happens automatically. Yet the pattern the scan hints at is a familiar one: roll out access to an AI platform, post a set of guidelines, and trust that classroom practice will adjust. That approach treats AI as a purchasing decision when it is closer to a pedagogical redesign.
The thinking in the field is moving the other way. Current work on AI literacy puts the weight on critical evaluation, ethical reasoning and sound judgment, not on button-pushing skills. UNESCO’s guidance on generative AI in education similarly calls for a human-centred approach that protects human agency, inclusion and accountability. If those are the goals, then investing in people, meaning the faculty and teaching staff who must interpret and apply the technology, deserves at least as much deliberate attention as investing in the technology itself.
Two things that get confused: literacy and credentials
The scan also surfaced a distinction that is easy to blur. Universities are busy launching AI certificates, microcredentials, degree programs and executive courses. Eight of the 15 institutions were earning revenue from some form of AI education.
Alongside that commercial activity, many of the same institutions publish free materials: library guides, responsible-use resources and other aids that help anyone on campus make sense of AI.
These serve different purposes. Credentials answer a labour-market question: how do we prepare graduates and working professionals for jobs shaped by AI? Literacy answers a broader civic and intellectual one: what should every member of a university community understand about AI in order to engage with it critically, responsibly and effectively?
An institution can be strong on one and thin on the other. A university with a polished portfolio of paid AI certificates might still offer little to the history major or the nursing student who simply needs to use these tools wisely. This split will matter more over time, as universities settle a pressing choice: is AI competence something for specialized programs, or an expectation of graduates in every discipline?
Size and centralization are not destiny
Perhaps the most counterintuitive finding concerns what did not predict success. Across the 15 universities, being large did not reliably translate into more initiatives, more AI research output or stronger entrepreneurial results, which the team gauged through disclosed projects, hackathons, incubators, startups and grants. Nor did a more centralized governance model produce better outcomes.
That should give pause to anyone assuming the answer is a bigger budget or a tighter chain of command. The study points instead to organizational readiness for change. Lasting transformation depends on whether the people inside an institution share a commitment to the change and believe they collectively have the capacity to carry it out. Who sits on which committee, and which office holds the mandate for AI, matters less than whether people across the organization actually buy in.
The university is not an island
The authors flag an important limit of their own method. Counting initiatives on a university’s website can mislead, because a university’s AI capacity does not stop at its own gates.
Canada’s three national AI institutes, Amii in Edmonton, Mila in Montréal and the Vector Institute in Toronto, knit together universities, researchers, governments and industry. Through them and through regional ecosystems, institutions gain access to expertise, research networks, talent pipelines, industry collaboration and routes to commercialization. Much of that value never appears on a campus web page.
Two universities with similar-looking public footprints could therefore have very different real strength, depending on the networks they plug into. A fair assessment has to treat these external relationships as part of an institution’s core capacity, not as peripheral extras.
Rethinking the “AI race”
Public conversation about AI often runs on the language of a race. Change is inevitable, the argument goes, so institutions must move fast, and every university is assumed to be running along the same track, with some simply ahead of others.
The scan complicates that picture. If institutions are pursuing genuinely different models, there is no single finish line, and being “behind” depends on which course you think you are running. Some universities are building toward research dominance, some toward commercial spin-offs, some toward teaching excellence, some toward coherent governance. These are different bets, not different positions on one leaderboard.
For leaders, that shifts the useful question from “how quickly are we adopting?” to “what are we adopting for?” The study offers a set of prompts that any institution could put to itself:
- Are students building the AI literacy and judgment they need to make informed decisions?
- Are instructors getting real support to redesign how they teach and assess?
- Is governance clear enough to allow experimentation while keeping risks in check?
- Do entrepreneurship and commercialization efforts connect to the university’s educational and public mission?
- Are outside partnerships treated as part of the institution’s own capacity?
- Taken together, do these pieces add up to a coherent approach?
No universal answer
The final point may be the most important one. The answers to those questions will not, and probably should not, be the same everywhere. A university rooted in a particular region, serving particular students and carrying a particular history will rightly arrive at a different AI strategy from one across the country.
What the scan ultimately argues for is intentionality. The universities best placed for the AI era are unlikely to be the biggest, the most centralized or the fastest to sign a vendor contract. They will be the ones that decide deliberately what AI is for within their own mission, invest in the people who must make it work, and build the networks and governance to match.
