Loading page…
The provisional scan covered 15 Canadian research-intensive universities. Its authors distinguish selling AI credentials from preparing an entire university community to use AI critically.
Listen to this story
The scan is a map of what 15 Canadian research universities publicly disclosed from March to June 2026, not a comprehensive audit of their AI work. Its authors say the gap between broad AI activity and initiatives explicitly supporting instructors points to a practical challenge: universities need to help teaching staff make decisions about assessment, responsible use and learning, not just provide tools or credentials. The provisional, un peer-reviewed findings also caution against ranking institutions by initiative counts, which miss differences in strategy and external connections.
Researchers counted 953 publicly documented AI initiatives; 91 explicitly targeted instructors’ teaching.
Eight of the 15 universities were monetizing some form of AI education, including certificates, short credentials, degrees or executive education.
The scan used university websites and keyword-based scraping, so it captures public documentation rather than all institutional activity.
Universities can offer AI credentials and tools without giving teaching staff the support to rethink learning. A scan described on October 12 found that just 91 of 953 publicly documented AI initiatives at 15 Canadian research-intensive universities explicitly targeted instructors’ teaching. The researchers argue that investment in people deserves as much deliberate attention as investment in AI.
Nadia Delanoy, an assistant professor at the University of Calgary, conducted the scan with Guy Levesque, executive director of the university’s Hunter Hub for Entrepreneurial Thinking. They described their findings in The Conversation. Their work examined publicly visible activity between March and June 2026, using university websites and keyword-based scraping.
The research has not been peer-reviewed, and the authors call it provisional rather than comprehensive. It maps what universities publicly document, not everything they do. The initiatives covered a wide range of activity: student learning, credentials, entrepreneurship, external collaboration, infrastructure, events, funding, student groups and governance, as well as teaching.
The teaching finding concerns initiatives explicitly aimed at instructors, not all education-related activity. The authors flag that distinction because instructors must decide when AI helps students learn and when it undermines the intended learning. They also face decisions about changing assessments, defining responsible use and preparing students before they hand tasks to an AI system.
Their concern is a response built around tools: give people access to a platform, publish guidelines and expect teaching practices to adjust. They argue that universities should invest deliberately in the human skills needed around AI. In their account, AI literacy means more than operating a system; it includes evaluating outputs, reasoning about ethics and making informed decisions.
Eight of the 15 institutions were monetizing some form of AI education. Universities were developing certificates, short-form credentials, degree programs and executive education. Alongside those offerings, institutions also provided free student resources, library guides and responsible-use materials. The scan therefore describes both paid education and broader support, rather than a simple choice between the two.
But the authors distinguish AI credentials from AI literacy. Credentials address preparation for a labor market shaped by AI. Literacy asks what everyone in a university community should understand to engage with the technology critically and responsibly. A university can be strong in one area while comparatively underdeveloped in the other, they argue.
That leaves an institutional choice: should AI competence sit mainly within specialized programs, or become something expected of graduates across disciplines? The authors frame that as a question of educational purpose, not simply a decision about which new credential to offer.
The scan found different institutional priorities, rather than a single adoption path. Some universities emphasized research and outside partnerships; others focused on entrepreneurship, central coordination or innovation spread across faculties. Still others concentrated on teaching, professional learning and AI literacy. Those choices reflected different views of what AI should do within a university’s mission.
Institutional size did not reliably predict initiative counts, AI research output or visible gains in entrepreneurship, the authors reported. Greater centralization of AI governance did not reliably predict those measures either. Their entrepreneurship measures included disclosed projects, hackathons, incubators, startups and grants—not just a university’s stated ambitions.
They also caution against comparing universities solely by counting initiatives. Connections through Canada’s Amii, Mila and Vector Institute can give universities access to expertise, researchers, governments and industry beyond what their own websites show. The authors urge leaders to treat those external relationships as part of institutional capacity, while asking whether commercialization efforts serve their educational and public missions.
Loading discussion...
Join the conversation
Explain whether this belongs in every degree or only specialized programs.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.