Editor’s note: This is the 16^th article since May 20, 2026 in an ongoing series by Dr. Andrew Maxwell, the Bergeron Chair in Technology Entrepreneurship in the Lassonde School of Engineering at York University. Every week – and occasionally every other week – we’ll present a new article by Maxwell, in a series whose wide-ranging and incisive themes encompass: Canada and innovation policy; productivity and industry; innovation frameworks; AI and higher education; research and intellectual property; technology adoption; entrepreneurship and commercialization; universities and higher education; entrepreneurship education; and AI and the future of work. 

This is Part 4 of a five-part series looking at the future of universities: Why are universities so difficult to change? What does AI mean for learning? If AI can teach, what is the professor for? And if some of the constraints around which universities were designed are disappearing, what might we design instead? Part 1 was published on August 26, 2026, Part 2 on September 2, and Part 3 on September 9.

Students settled it before most institutions finished writing their first policy.

They are using AI to summarize readings, test ideas, draft outlines, debug code, prepare presentations and ask the questions they may not feel comfortable asking in class. Faculty may still be debating whether this is good or bad, but the practical reality is clear: AI is already part of the learning environment.

So the real question is no longer whether AI will enter higher education.

It already has.

The real question is whether universities will use AI to improve learning, or whether we will simply bolt it onto old educational structures and hope for the best.

That distinction matters.

If AI is treated mainly as a shortcut, it may weaken learning. Students may produce more polished work without becoming more capable. They may outsource thinking, hide behind fluent writing, and confuse completion with understanding.

But if AI is designed into education with a clear learning purpose, it can do something much more valuable. It can help more students access the kinds of guidance, feedback, reflection and coaching that were previously scarce. That is the real case for AI-enhanced learning.

For decades, we have known that students need more than content knowledge. They need to learn how to frame problems, reason with evidence, communicate, collaborate, make ethical decisions, use judgment and act under uncertainty. These capabilities are difficult to develop through lectures and final assignments alone. They require practice, feedback, reflection, iteration, and support.

The problem has always been scale.

A faculty member can coach a small group of students. A mentor can help one learner make sense of a difficult project. A teaching assistant can support a team. But providing timely, personalized, process-oriented guidance across large courses, multiple programs, and diverse learner pathways has always been difficult.

AI changes that capacity question.

Not because it replaces professors.

Not because it writes better assignments.

Not because it makes education cheaper.

AI matters because it can help make high-quality learning support more available. It can prompt students to clarify assumptions, ask better questions, interpret feedback, reflect on their progress, and prepare more thoughtfully for human interaction with faculty, peers, mentors, and employers.

This is especially important because many students do not struggle simply because they lack ability. They struggle because the hidden processes of learning are often unclear. They may not know how to begin an open-ended assignment, how to improve after feedback, how to evaluate alternatives, or how to connect one learning experience to another.

A well-designed AI learning agent can make some of that hidden curriculum visible.

It can ask the question a student may not yet know how to ask. It can challenge vague reasoning. It can help a learner see that their first answer is not the end of the work, but the beginning of deeper thinking.

That does not make learning easier.

In many cases, it makes learning more demanding.

A good AI learning agent should not simply say, “Here is the answer.” It should ask, “What evidence do you have?” “What assumption are you making?” “How did feedback change your thinking?” “What would you do differently next time?”

That is not answer generation. That is learning support.

Shifting from transactional learning to developmental learning

The most important shift is from transactional learning to developmental learning. In a transactional model, students complete courses, submit assignments, receive grades and accumulate credits. Those things still matter. But they do not fully capture what learners need in an AI-shaped world.

A developmental model asks a deeper question: how is the learner becoming more capable?

This is where AI-supported reflection becomes powerful. Students often complete projects without recognizing how they have grown. They may receive a grade but miss the deeper evidence of development: that they became better at asking questions, working with others, handling uncertainty, revising their thinking, or taking ownership of their learning.

AI can help make that growth visible.

It can help students build evidence of capability over time, not just evidence of completed assignments. That matters for employers, graduate programs, professional identity and the student’s own confidence.

Of course, this will not happen automatically.

AI adoption in universities should not be judged by whether institutions appear modern, efficient or technologically sophisticated. It should be judged by whether AI strengthens the learner experience and helps students develop capabilities they could not develop as effectively through existing models alone.

That is the case for AI-enhanced student learning.

Not AI as a shortcut.

Not AI as a replacement for faculty.

Not AI as educational theatre.

AI as a way to make learning more guided, reflective, developmental, and visible.

The danger is not that AI will make students too capable.

The danger is that it will let them look capable without becoming capable.

That is the challenge universities now face.

R$