Will AI Change The Way We Build Universities?

Almost everyone who teaches or studies in higher education will recognize this scene. You walk into a brand-new lecture theater, put your laptop on the desk and discover that you have 14% battery left. What follows is an activity that was never included in the curriculum: the hunt for a power socket!

In some institutions it is under the seat. In others it runs along the wall, which forces students who arrived late to negotiate with their neighbors. And in a few lecture theaters, despite considerable investment in digital transformation, nobody seems to have thought students might need one.

The anecdote is amusing, and it tells the recent story of university campuses rather well. For the past 20 years, higher education institutions have been chasing technological change and trying to translate it into their buildings, while their students carry extension cords around in their bags.

First, campuses had to become connected. Not so long ago, having Wi-Fi everywhere was still worth a mention on a campus tour. Today, asking a student to choose an institution for its excellent Wi-Fi would sound about as strange as proudly announcing that the buildings have running water.

Then came laptops, and therefore power sockets. Next came giant screens, projectors, interactive whiteboards, videoconferencing systems and hybrid classrooms. The pandemic sped everything up: within months, institutions had turned rooms designed for one professor and thirty students into small television studios.

With each new wave we added equipment to the classroom and left its layout alone. After 20 years of innovation, most rooms still look like the ones our parents studied in, with a professor at the front, students in rows, and a space designed so that one person can speak to many. A campus can show off the latest equipment and still run on a very old idea of learning, and of who does what in the room. We have sometimes installed hundreds of thousands of euros’ worth of technology without asking what we wanted to happen there.

Artificial intelligence may be a different kind of disruption. Screens and sockets stayed on the walls and the desks, whereas AI takes part in the lesson. As Alain told a room of university leaders in Paris in March 2026: AI is no longer in the corridor. It is in the room!

FROM THE METAVERSE TO A HEADSET ON A PLANE

Higher education has already been through several technologies that were supposed to reinvent the campus. A few years ago it was the metaverse. Some universities bought virtual land, built digital campuses and organized classes where students and faculty met as avatars. For a while, you could almost imagine that a university’s next major real estate investment would need neither an architect nor planning permission.

In September 2020, NEOMA Business School opened a persistent virtual campus in addition of its three sites in France (Reims, Rouen and Paris), in a version that ran on an ordinary laptop (with and without a headset), so that no student was left out. Parts of it worked well: students sat through three-hour classes there without the fatigue that video calls produce. It was built to complement the physical campuses, never to replace them, and the wider enthusiasm for virtual land has since faded due to the deep need of human interactions, in the real life.

Virtual reality has not disappeared. It is finding its place in narrower uses: technical training, medicine, engineering, simulation, and access to places that would otherwise be hard to visit, and NEOMA still invests in this technology.

Air travel shows how difficult these bets are to call. In 2017, American Airlines began removing the screens from its seatbacks, on the grounds that most passengers carried a phone or a tablet. In August 2026 it announced that it would put a screen back on every narrowbody seat, starting in 2028. Passengers, it turned out, were using their own devices for work and messages. Lufthansa, meanwhile, has been running a pilot since 2024 that lends mixed-reality headsets to some business-class passengers. In less than ten years, one industry has taken screens out, decided to put them back, and started testing headsets!

Classrooms face the same uncertainty. Fifty open laptops in a lecture theatre is normal today; twenty-five years ago it would have looked like a particularly strange computer lab. Mixed-reality glasses may one day seem just as ordinary, though we have heard enough predictions about the death of the computer to stay cautious, and three hours with a headset strapped to your head, with the motion sickness some people get, is probably not the future most students dream of, for now!

WHEN THE VIRTUAL CEO STARTS ANSWERING BACK

Virtual reality could already take a student into a factory, a laboratory, or a boardroom. For a long time, though, the experience stayed largely scripted. Students could look around and perhaps make a few choices, and the characters they met followed a predetermined scenario.

With generative AI, the characters answer. A strategy student can walk into the office of a virtual CEO and ask why sales are declining in Asia. The CEO replies. The student can challenge the response, ask for the numbers, or go back to a decision made three years earlier. The CEO might become impatient, refuse to answer, or give away a piece of information that sends the student down a different path. A few metres away, another student meets the same CEO and has a completely different conversation.

This is already in use. At NEOMA, every first-year student in the Grande École Programme (1,000 students) now practices commercial negotiation against an AI simulator, across several business scenarios, as part of a negotiation course. Poets&Quants for Undergrads also reported in September that more than 2,000 students at Northwestern’s Kellogg School have used case studies rebuilt as simulations with AI-generated characters.

Immersive technology supplied the setting, and AI now supplies the people in it. That changes what we expect from a classroom, and possibly what we call one.

THE CLASSROOM COULD BECOME AN EXPERIENCE ROOM, AND THE EXPERIENCE COULD BECOME THE CAMPUS

Another scene will be familiar to many teachers. You walk into a so-called “flexible classroom” and spend the first ten minutes moving tables that were supposedly designed to encourage collaborative learning. Active learning begins with a collective furniture-removal exercise.

Research that one of us co-authored on store design found that how easily shoppers can move around a space counts about as much in their experience as its atmosphere does. We see little reason to think classrooms are an exception!

If we want more experiential learning, we should design rooms around what students are supposed to do in them, and choose the technology afterwards.

A future experiential learning room might be far less spectacular than today’s smart classroom. It would have tables that are easy to move, small spaces for group work, excellent acoustics, the possibility of some privacy, reliable connectivity, a few screens where they are useful, and plenty of power sockets, because there is no need to be too futuristic.

It is often these ordinary details that change a student’s day: being able to hear one another, finding somewhere to work together, having a place where students want to stay after class.

Behind that simplicity sits a digital infrastructure that can change what the room is for. For one hour it is the headquarters of a company facing a crisis. The next, several teams are negotiating an acquisition, and in the afternoon students appear before a virtual investment committee. Through all three, the room itself barely changes.

A room that records students negotiating, hesitating and losing their temper is also collecting sensitive data. Someone has to decide where that data is kept and who gets to see it. In Europe, it is barely forbidden due to GPDR & AI Act, but it is allowed in other places.

THE CAMPUS AS A SIMULATOR

Nobody would hand over an Airbus to someone who had only attended excellent lectures on aeronautics. Pilots spend hours in simulators because some skills cannot be acquired by listening to someone explain them. You have to make a decision, see what happens, make a mistake and try again.

Is management so different? We explain negotiation to students before they have negotiated, teach them leadership before they have managed a team, and analyze crises they have never lived through. Between two classes there is also everything that no curriculum can plan for: convincing a classmate, managing frustration, finding your place within a group.

Case studies were developed to reduce this distance between knowledge and decision-making. AI goes one step further and puts students inside the situation they used to analyze from the outside.

A future manager can announce a restructuring to a virtual employee and discover that the way the message is framed changes the whole conversation. The same goes for a finance student defending an acquisition before a committee that challenges every assumption, or a marketing student presenting a campaign to a skeptical CEO. All of them can try again, and failure becomes one more step in learning.

Pilots do not train alone, though. An instructor sets up the failure, watches how the crew handles it and takes them through the session afterward. That is the professor’s job in an experience room: less time delivering content, more time as a sparring partner who challenges students and makes them justify their choices.

WHAT IF AI MADE THE CAMPUS MORE IMPORTANT?

On the campuses we visit, students rarely talk first about screens or equipment. They talk about a person they met or a project they managed to see through. It may sound like a minor detail of student life, yet this is often where a sense of belonging to an institution comes from.

We have spent a great deal of time describing the campus as a place for transmitting knowledge. It is also where students meet people from other disciplines, find their allies, learn to speak up and learn to make room for someone who does not think or look like them.

AI can help a student prepare for an interview or understand a difficult concept at midnight. It cannot create the conversation that continues after the class has ended, or the trust a group of students needs before they decide to build something together.

Since generative AI arrived, we have been asking whether students will still need to come to campus. If everyone has a personal tutor that can explain any concept, in any language, at any time, then part of what we used to call “going to class” can happen somewhere else.

The evidence on learning alone with AI is less comfortable. A working paper by economists at Stockholm University and the University of Hong Kong followed 26,811 Chinese secondary school pupils for 30 months. Six months after pupils started using generative AI, their homework marks had risen by 18% and their exam scores had fallen by 20%. Secondary school in China is a long way from a business school, the authors could not see how pupils used the tool, and other studies have found that AI tutors designed for teaching improve learning. Still, many teachers will recognize the pattern: work done alone at home tells us less and less about what a student has learned. Learning needs friction, effort, confronting to others and an AI that does the work removes it. On campus, the reasoning can be seen, explained, reversed, by a professor and by other students.

So the easier it becomes to learn alone, the more it matters to create reasons to come and learn together. The campus can be the place where students negotiate, debate, test decisions, fail, try again and have a professor help them understand why.

If homework no longer shows what a student knows, more assessment will move to oral defenses and to work done in the room, and buildings will have to follow. That takes many small rooms, fewer large theaters and a great deal of faculty time, and to our knowledge nobody has yet put a figure on the cost.

University buildings also last much longer than the technologies inside them. A classroom built today will probably still be in use in 2050 or 2060, by which time its digital tools will have been replaced several times. Designing tomorrow’s campus around today’s technology may therefore be a dangerous approach without flexibility.

We have moved from Wi-Fi to power sockets, from projectors to smart classrooms, and from the virtual reality to the metaverse. Tomorrow we may all wear glasses or headsets instead of using computers. Or, like American Airlines, we may change our minds!

Whatever device students end up with in front of their eyes, we can decide now what we want them to do once they have made the effort to come to campus. For a long time, we built classrooms to transmit knowledge. Now that AI makes some of that knowledge available everywhere and instantly, we should build more rooms in which students can experience it.

And, as a starting point, we could simply make sure there are enough power sockets!

---

Benjamin Stevenin is the former Director of Business School Solutions and Partnerships at Times Higher Education and the founder of Deans Insight, a new research platform that tracks the career paths of business school alumni to give schools data-driven insight into their graduates’ outcomes. Alain Goudey is a marketing professor and Associate Dean for Digital at NEOMA Business School in France. Julie Giraud-Avril is the founder of Conseil x JGA, a consulting practice specializing in higher education, innovative learning ecosystems, and campus development.

© Copyright 2026 Poets & Quants. All rights reserved. This article may not be republished, rewritten or otherwise distributed without written permission. To reprint or license this article or any content from Poets & Quants, please submit your request HERE.