Teaching Finance in the Age of AI

In November 2025, Professor Kerry Back received the Financial Management Association’s Innovation in Teaching Award. 

We sat down with him to talk about how AI is changing finance — and the way we teach it.

A lot of people are feeling anxiety about AI. You’ve written about how this isn’t the first time education has gone through a shift like this. How do you think about that history as you’re teaching in the finance classroom?

It’s worth remembering that every major technology has created classroom anxiety. When calculators came out, people worried students would never learn math. When Excel showed up, there was panic about students losing the ability to compute by hand. Socrates famously worried that writing would “produce forgetfulness” and create “the appearance of wisdom, not true wisdom.”

Generative AI fits into a very old pattern. The tools change, but the core concepts don’t. Students still need to understand finance. They still need to interpret results, question assumptions and think critically. 

What changes is how much time we spend on the mechanics. We don’t teach square roots by hand anymore, and that didn’t hurt anybody’s understanding of math. AI is simply another step in that direction.

You started trying AI tools with students early on. Where did that interest begin?

Honestly, it started with realizing how much we were still forcing into Excel because it was the tool we had. Spreadsheets are perfect for pro formas and discounted cash flows, but they’ve never been great for a lot of the other analyses people in industry now handle with more powerful tools.

Once the AI models became good enough to write and run code, the barrier disappeared for students who don’t know how to program. Suddenly, students could use the same kinds of tools professionals rely on without being programmers. That’s what pulled me in. 

As the tools have improved, my teaching has shifted naturally from showing students what AI can accomplish to having them build with it: apps, workflows, automations, etc.

What does AI-enabled teaching look like when you’re in front of students?

The biggest shift is that we treat AI as the interface. Instead of opening Excel or searching through menus, students “chat” with the model. They tell it to pull data, sort stocks or run regressions — and the model writes the code and executes it.

I’ve even built what are essentially “skills” for the model — prompts that tell it exactly how to perform certain finance tasks — so students can say, “Use your skill,” and it knows what to do. 

The goal isn’t to turn students into coders. Instead, they’re designing workflows, checking results and iterating with the model until the analysis is correct.

When students first work with these tools, what jumps out at them?

The first thing that hits them is just how much AI can do. Most of our students have never programmed, so watching the model generate code feels almost unreal. But then they hit the other side. They’ll ask the model to do something, and it gives an answer that doesn't make sense.

My classes help them shift their mindset. You have to chat with the tool like a colleague until you get the result you want. Once they get past that, it gets exciting. Things they once struggled to build in spreadsheets can be automated with a simple prompt. I had a student say, “I’ve never programmed before, and now I can build apps.” Moments like that tell you it’s working.

How closely does that classroom work track with what finance firms are doing right now?

Pretty closely. Firms are trying to automate the exact same things students practice — the repetitive analyses and formatting that eat up so much time for junior analysts. 

What firms really need are people who understand finance and can also work with AI to build simple, customized tools. In some ways, the classroom is actually ahead of the curve, because students can experiment without the constraints firms face in production.