Ivey Dean Julian Birkinshaw: Don’t Expect A ‘Killer App’ For AI In B-Schools

Julian Birkinshaw, dean of the Ivey Business School at Western University, has reviewed fourteen AI tools built for business school classrooms, and he’s ready to disappoint the edtech companies banking on one of them becoming indispensable.

“I think the edtech companies that are trying to get into this space are going to be disappointed,” Birkinshaw tells Poets&Quants in an interview about his new white paper, From Chalkboard to Chatbot: A Review of AI-Enabled Learning Tools for Business Schools. “We’ve kind of seen this movie before with online learning.”

Birkinshaw, who co-authored the report with Ivey assistant professor Mazi Raz and Markus Walters, director of strategic initiatives at the London, Ontario B-school, says the barriers to building these tools are so low that no single product is likely to dominate. “Faculty at Ivey, and probably at other places as well, can in a couple of days knock up a little AI learning bot which does 80% of what some of these really expensive things do,” Birkinshaw says.

A MICROSOFT MOMENT, NOT A KILLER APP

Rather than one breakout winner, Birkinshaw expects AI features to get absorbed into the platforms schools already use. 

“If you look at Canvas and Moodle, these various LMSes, it’s just like Microsoft, isn’t it,” he says. “You basically take whatever the hot new app is and just fold it into what you’re offering.” Specialist tools will survive alongside that, he adds, but he doesn’t expect any of the fourteen tools in his survey, seven built at Ivey and seven external, to “take over the world.”

He also draws a direct parallel to the MOOC boom of the early 2010s. 

“It was around 2012 when that all kicked off,” he says. “For about five years they were booming, and then COVID proved that you could indeed do learning online, and then we figured out we didn’t want to do learning online.” He sees a similar cycle ahead for AI tools, though he’s careful to note the stakes are higher this time around, since the issue isn’t delivery format but “a much more fundamental issue of how students learn.”

THE NOVELTY PROBLEM

Birkinshaw offers a specific, sobering example from his own classroom. He built AIBEL, a voice-based tutoring tool, and ran a controlled test: some students were randomly assigned to use it before a case discussion, others were not. The first time students used AIBEL, their in-class participation jumped noticeably. Then the effect disappeared.

“That’s kind of a classic Hawthorne effect,” he says, referring to the tendency for people to change behavior simply because they know they’re being observed or trying something new. “The real risk is that you take it easy, you find ways of not using it in a serious way.”

That, he says, is the central problem facing every tool in his survey. Make a tool optional and many students will simply skip it. Mandate it through a grade and some students will find workarounds, including building their own AI bot to interact with the faculty’s AI bot on their behalf. 

“That’s not a good outcome either,” Birkinshaw says. “I don’t think anyone’s got the answer to that.”

His tentative fix leans on metacognition: teaching students to think about how they learn best, rather than simply handing them a tool and hoping they engage with it honestly.

LET A THOUSAND FLOWERS BLOOM. THEN WHAT?

Several of the strongest tools in the report, including AIBEL, Sidekick and Hopeworks, were built by individual Ivey faculty with little institutional backing. Birkinshaw calls that decentralized approach a deliberate first stage, not a permanent strategy.

“Some of those individual faculty will undoubtedly let those ideas slide, and that’s just human nature,” he says. The next step, he says, is figuring out how to invest modestly in the tools that show promise so they outlive the professor who built them. “I don’t expect anybody to make a whole lot of money out of this,” he says. “But making it easy to use, so that somebody who’s not quite so techy can pick up on this thing and fold it into their course development, that’s the goal.”

Asked whether schools should be formalizing that kind of support, he doesn’t hesitate. “Yeah, I think that’s exactly right,” he says. “Schools should be trying to find ways of either buying or building simple tools which give structure to learning that students benefit from, so faculty can buy into them at minimum cost.”

WHERE AI SHOULD BE BANNED

The report argues for explicitly AI-free zones in the classroom, and Birkinshaw is unambiguous about where that line sits for assessment. “In terms of assessments, there’s no argument here,” he says. “You have to assess using closed-book exams or oral exams, otherwise students will cheat. I don’t think that’s controversial.”

The harder question is the live case classroom, a format central to Ivey’s pedagogy. 

“Many of my colleagues now enforce a policy where they say laptops down, phones checked at the door, and we’re going to have a 40-minute discussion without anybody having any access to their devices,” he says. 

That forces students back into articulating and defending a position out loud, he says, though he’s not fully confident it will be enough on its own. 

“I can imagine many schools will struggle to even enforce that level of banning AI. For me, at the moment, that’s the best-case scenario I’ve got.”

NOT WORRIED ABOUT AN EQUITY GAP, YET

With tools like Lexarius and CAiSEY built in partnership with well-resourced schools such as INSEAD, Wharton and Columbia, while Ivey’s own AIBEL remains an early-stage prototype, one might expect Birkinshaw to worry about a widening gap between schools that can afford sophisticated AI tools and those that can’t. 

He isn’t, largely because he doesn’t believe a killer app is coming that would make that gap decisive.

“Some of these things will be done on literally $10,000 budgets rather than $10 million budgets,” he says. 

He argues some well-funded edtech companies have moved too fast toward polish before the underlying pedagogical problem is fully understood. 

“I’m not going to point a finger at any one of them, but some of them have jumped the gun,” he says. “It’s actually a bit of a waste of money to create a polished product when we aren’t quite sure what the real benefits of AI in learning are.”

3 THINGS THAT ACTUALLY WORK

Pressed on where AI genuinely helps rather than hurts, Birkinshaw points to three patterns from his survey. 

First, tools that act as a “traffic cop,” slowing students down and forcing them through a structured process, as New Venture Development does. Second, tools that act as a “sparring partner,” where the AI asks questions and the student has to answer them, rather than the reverse. 

And third, tools that make student learning visible to faculty.

He singles out CAiSEY, built by Columbia’s Dan Wang, as the clearest example of that third benefit. 

“The real benefit I’ve got out of creating this tool is I can now see everything they’ve learned and everything they’ve not learned,” Birkinshaw says, paraphrasing Wang. “That helps him pitch his classes at the right level, on the cusp of their knowledge, rather than either going too far or covering old ground.”

‘WE ARE STILL IN THE FOOTHILLS’

Birkinshaw is blunt about how early this all still is. “We are still in the foothills,” he says. “Everyone’s experimenting. Nobody knows quite what the right way forward is.” He expects business schools to eventually move from individual faculty tinkering in isolation toward a more systematic rethinking of entire courses and programs, pointing to Ivey’s New Venture Development project as the closest example so far of a course reshaped end to end around AI.

He also draws a comparison to how the MOOC wave eventually gave way to smaller, more focused offerings from companies like Emeritus and 2U, sometimes called the SPOC model, blending online delivery with in-person teaching. “MOOCs haven’t died, but MOOCs have just become the low end, anything’s available for free,” he says. “As long as you don’t get any kind of credit from it.”

As for predictions that edtech startups might disrupt business schools outright, Birkinshaw is confident they won’t. 

“Our business model is only in part about learning,” he says. “It’s about the credentials, the networks, the accreditation. As long as we stay on our game and adapt and adopt these new technologies, this will rapidly become what Christensen calls a sustaining technology, not a disruptive one.”

DON’T MISS THE P&Q INTERVIEW: IVEY DEAN JULIAN BIRKINSHAW HAS GLOBAL AMBITIONS FOR THE ELITE CANADIAN B-SCHOOL

© 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.