AACSB Asks: In The Race To Teach AI, Are Students Still Learning To Think?

Consider a classic business-school assignment: A student submits a polished analysis complete with a coherent argument and several persuasive recommendations. A deliverable that is both client and instructor ready.

In the before times, producing that work required research, weighing evidence and competing incentives, and making decisions on what mattered most. The struggle was part of the lesson.

Today, AI can produce a similarly polished result in a fraction of the time, with far less mental effort. But what, exactly, did the student learn? And what should the business professor evaluate?

“Having the answer to everything all the time is not necessarily the best for learning or development,” Steven D. Shaw, a visiting scholar at the Wharton School, told a room of business school deans at the AACSB conference last week in Orlando.

Shaw, an assistant professor at King’s Business School at King’s College London, studies how AI reshapes human reasoning and decision-making. He was one of three panelists for “The Cognitive Cost of AI,” among several sessions examining how business schools are responding to a technology reshaping both education and the work graduates will do.

The discussion exposed a tension at the heart of that response: Schools must prepare students to work with AI without allowing it to replace the thinking that prepares them to work independently.

For deans under pressure to integrate AI across their programs, that means looking beyond whether students can produce better work with the technology. They also need to know whether students are developing the knowledge and judgment to recognize when that work is wrong.

DEANS’ AI AMBITIONS OUTPACE SCHOOL AI PREPAREDNESS

AACSB’s Deans Conference brought together about 700 business school leaders and other participants to examine the pressures reshaping management education. AI ran through the agenda, from sessions on curriculum and governance to discussions of career readiness and the future of the MBA.

In conjunction with the conference, AACSB released its 2026-27 Deans Survey, which it conducts every three years. Among respondents, 87% of B-School deans rate AI readiness a high or very high strategic priority, but only about one in five give their schools similarly high marks for preparedness.

The conference offered several perspectives on what closing that gap should involve. Researchers emphasized independent thinking and the effort required to develop expertise. Former Amway CEO Milind Pant urged schools to build more learning around customers, real business problems, and leading change. AACSB President and CEO Lily Bi argued that the AI transition should elevate the human dimensions of education, including mentoring and personal development.

Together, their remarks raised a question more demanding than whether business schools are adopting AI quickly enough: What should students still have to learn the hard way?

DOES AI LEAD TO ‘COGNITIVE SURRENDER’?

On the “Cognitive Cost of AI” panel, Shaw joined Purdue University associate professor Brooke N. Macnamara and Pant to examine what students risk losing when they delegate thinking to technology.

Shaw distinguished between cognitive offloading — using a calculator or writing down a number — and what he calls “cognitive surrender.” Generative AI can frame a problem, supply the reasoning, and deliver a conclusion, allowing users to surrender oversight of the process itself.

People already seek ways to reduce the effort of thinking, he said. AI makes those shortcuts readily available.

“Having the answer to everything all the time is not necessarily the best for learning or development,” Shaw said.

Not all offloading is harmful, said Macnamara, an associate professor of psychological sciences at Purdue who studies how people acquire expertise and how AI use might contribute to skill decay among professionals.

The question is which abilities students need to develop. If the assignment calls for an attractive PowerPoint presentation, judging the finished product may be appropriate. But if the goal is integrating information or developing an original research question, instructors need evidence of students’ own capabilities.

“What skills are they developing independent of AI?” she asked. “And in that case, we need to measure it independent of AI.”

That distinction is especially important for beginners. Experienced faculty and managers may use AI to support skills developed over years. Students are still building that foundation.

Macnamara pointed to “desirable difficulties”: effortful activities, such as testing yourself rather than rereading, that help build and retain knowledge. AI can remove that effort, leaving students feeling productive without doing the thinking themselves.

Her advice: Attempt the problem or write the essay first, then use AI for feedback, practice questions, or help identifying weaknesses. In other words, think first and then write prompts to identify gaps and improve your work.

Schools also need to examine what they reward. Grading polished output alone may encourage students to delegate reasoning. Shaw suggested assessments without AI to check whether students can demonstrate the intended skills.

WHAT EMPLOYERS WILL STILL NEED

From the employer perspective, Pant presented a more optimistic view for business students. He sees the first AI-native generation entering work with the most powerful tools for tackling business and societal problems at their fingertips.

But access to those tools will not, by itself, distinguish exceptional leaders.

“Exceptional business leaders always start with the customer,” Pant said. “They start with where the rubber hits the road,” Pant said.

He emphasized three skills business schools must teach: understanding customers through firsthand experience, connecting information across contexts, and bringing people together to lead change.

He pointed to his own career as an example. After business school, he joined Unilever expecting something closer to a conventional marketing role. Instead, he found himself working on supply-chain and vendor problems in leather tanneries in small Indian towns.

The smell has stayed with him for 35 years, he joked, but so have the lessons. Working with colleagues, he learned about manufacturing, maintenance, labor scheduling, quality, customers, and cash flow. It is still the foundation that he uses in decisions and judgement he makes today.

He urged schools to embed practical learning throughout their programs, with “not just one summer internship, but real world experience built and completely diffused throughout it.”

Pant also questioned whether traditional academic divisions provide the best structure for preparing students to solve problems that cross them. Future programs, could be organized more deliberately around major business and societal challenges, bringing multiple disciplines to bear, he argued.

While AI may make analysis more accessible, leaders will still need to understand the circumstances in which it will be used and persuade people to act on it.

BI’S CASE FOR ‘KNOWING, THINKING & BEING’

In her opening remarks, Bi described the educational task through knowing, thinking, and being.

Even if information becomes freely available, students need knowledge as a foundation for judgment. They need to understand enough to frame questions, supply context, and evaluate what AI produces. That’s what business schools must be prepared to teach.

B-schools need to pay more attention to the development of the person. Employers are looking now for perspective, resilience, and judgment alongside the hard skills.

“Research is great, teaching is great, but mentoring and inspiring is important,” Bi told the deans.

In a subsequent interview with Poets&Quants, she described an approach she had heard about from schools in Korea: dividing coursework into portions where students must use AI and portions where they cannot.

The idea is to prepare students to use the technology while retaining the foundational knowledge and skills they may need independently.

AACSB’s broader ambition is to share approaches schools can adapt to their own circumstances, including different models for responding to technological and financial pressures. Bi wants schools to see that work as an opportunity to experiment, rather than assume every new idea from AACSB signals an accreditation requirement.

“We allow exploration. We allow different varieties. We allow diverse business model, new things,” Bi told the deans in the conference ballroom.

For Bi, the uncertainty surrounding AI makes the foundations of education more essential.

“If a student doesn’t have knowledge, where is their foundation for judgment?” she asked. “Where is the foundation for thinking?”

DON’T MISS: U.S. B-SCHOOL APPS PLUNGE 25% AS GLOBAL DEMAND SHIFTS TO EUROPE & ASIA AND ‘NOW IS NOT THE TIME TO RUN AWAY!’: EUROPE’S DEANS RESPOND TO THE CALL TO ABOLISH B-SCHOOLS 

© 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.