How Duke’s Fuqua Is Using AI To Transform The Business Classroom

Duke’s Scott Dyreng:

AI is the biggest challenge to education in several generations. Universities and K–12 schools alike are grappling with AI’s disruptive potential to turn students into “prompt monsters” who can’t write an essay on their own or hold a basic conversation without a chatbot’s help. The prevailing view among parents, students, and educators is that AI is a threat to learning and to fair assessment of skills. In many places, its use in education has been limited to automating grading or outsourcing faculty office hours to chatbots.

At Fuqua, we are flipping the script. We began with the question “Can we use AI to strengthen live, in-person discussion rather than replacing it?” The goal is to turn classrooms into dynamic ecosystems where technology amplifies human connections, and where both professors and students receive rich, objective feedback aligned with Fuqua’s core values: consequential leadership, the collaborative spirit known as Team Fuqua, and the balanced development of intelligence, emotional strength, and decency.

Here is how the technology works. We start with the recording setup already in place in our classrooms, then expand the network of microphones and cameras to capture conversations more accurately. Students check in to their seats using an internally developed app, so the system knows exactly who is speaking. Custom software matches each spoken comment to a seat and produces a time-stamped transcript with accurate attribution of each utterance. In real classes, early tests already show accuracy above 90%, and refinements are underway to support different teaching styles, vocal tones, and room layouts.

A guiding principle in building this system has been privacy and ethics. Students in the beta phase opt in. Importantly, there is no change to student privacy or to our existing recording practices: the raw classroom recordings already exist. The breakthrough is that we can now automate and analyze data that was previously too costly or time-consuming to use. The output is available only to the professor and is used solely for teaching and learning.

The most obvious application is better assessment of attendance and participation. Attendance becomes automatic and exact. Participation grading moves from rough impressions to detailed metrics. The system can flag whether a comment was relevant, showed genuine understanding, or prompted others to speak. Professors will soon see which students spark the widest and most diverse conversations and which comments consistently shut them down. That data is like giving every professor the equivalent of a full-time teaching assistant in every class.

AI has advantages over human assistants in several important ways. It ignores appearances and accents. It does not get tired or distracted. It can spot patterns that people miss like the professor who unintentionally calls on the right side of the room much more than the left. The result will be fairer feedback to students and faculty and a deeper understanding of classroom dynamics.

Students will receive individualized reports on their contributions over time. If a student’s participation has been weak in recent sessions, the system can generate prompts that encourage more thorough preparation for upcoming classes. Over time, that feedback loop nudges students to come to class ready to engage, not just to sit and listen.

The larger opportunity is for faculty. Our professors are world-class researchers and teachers, but traditional student evaluations offer limited guidance on how to improve teaching effectiveness. This system changes that. Professors can see which topics consistently confuse students, which cases generate the most insightful debates, and which questions produce the strongest engagement. They can connect patterns in classroom discussion to exam performance and course evaluations to pinpoint what truly works.

The system will also give us a new way to see Team Fuqua in action. We will be able to measure which exercises draw in the greatest number of students and which ones elicit the widest range of viewpoints. Those broad patterns are almost impossible for a professor to track in real time while managing the flow of discussion—but they are precisely the kind of pattern recognition at which AI excels.

We are not waiting for AI to destroy education, nor are we waiting for Silicon Valley vendors to dictate how AI should be integrated in our curriculum. Instead, we are leading an effort to make AI transformative and collaborative, interacting seamlessly with teaching traditions and the values that have defined our school for decades. This would not be possible without the commitment of the faculty and students or the engagement of our wider community, including Anton Sahazizian (MBA ’00) and his wife Laura, who generously supported this initiative. Early tests are promising. As feedback from the AI classroom rolls in, faculty members are already imagining new ways to teach and new kinds of classes to design.

Business education needs to show that AI technology is not something to be feared, nor does it have to replace face-to-face discussion. Used thoughtfully, it can make human connection more meaningful and human judgment sharper. That is what Fuqua is building, one classroom at a time. While the full version of this system is still a few months away, the direction is clear. When fully deployed, every discussion at Fuqua will be a little smarter, a little fairer, and a lot more consequential.

Scott Dyreng is the Senior Associate Dean of Innovation and Professor of Accounting at Duke University. His research interests are in corporate tax avoidance, international taxation, and accounting for income taxes. He teaches managerial accounting to graduate students, and has received the Excellence in Teaching Award in the Duke MMS program three times.

© 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.