A Business School Professor's Lessons From Teaching AI To MBAs

August 25, 2026 - 17:33

After a full semester of teaching MBA candidates how to manage AI agents, Ayse Ozturk walked away with a clear conclusion: the thing business schools are not teaching is judgment. Not technical fluency, not prompt engineering, not even strategy. Judgment.

Ozturk, a business school professor, spent the term pushing her students past the usual AI coursework. They learned how to build simple agents, how to evaluate model outputs, and how to think about automation in a corporate setting. But what struck her was how quickly the students latched onto what the tools could do, and how rarely they asked whether they should.

The real gap, she says, is the ability to decide when to trust an AI output, when to override it, and when to throw it out entirely. That is not a technical skill. It is a human one. And it is getting harder to teach because the tools keep changing.

Her students were sharp. They could spot a hallucinated statistic or a biased recommendation when they were looking for one. But in messy, real-world scenarios, where the data is incomplete and the stakes are uneven, they defaulted to treating the model as a reliable partner. That instinct, she argues, is dangerous.

The lesson she wants to leave them with is simple: AI can draft, analyze, and even decide. But someone has to own the call. That someone has to understand the context, the tradeoffs, and the people affected. No model can do that yet.

Ozturk is not calling for less AI in the classroom. She wants more focus on the messy parts of management that machines cannot simulate. Because in ten years, her students will not be judged by how well they use AI. They will be judged by what they do when AI is wrong.