AI, Data, and the Execution Gap in Banking

Hey everybody. Welcome back to another episode of What’s Going On in Banking. I’m here, of course, with my co-host Stacey Bryant.

Stacey, how are you doing?

I’m great, Ron. My voice is a little rough too, but how are you doing?

We just got back from Phoenix, Arizona, for Bank Director’s Acquire or Be Acquired conference. It was such a pleasure to be there with you.

I’m doing great. It’s good to be back.

We’re recording this on a Thursday, and in two days I’m heading back to the Phoenix area. I have a few things to do at the office next week, and this time it’ll be nice to give my wife a break from the weather here in the Northeast.

It has been brutal.

For everybody who lives somewhere warm, be thankful.

But let’s skip the chitchat today because a lot happened.

First, I want to give a huge shout-out to Bank Director.

Acquire or Be Acquired is one of the best-run conferences in the industry.

The location is great, the organization is strong, and the quality of the crowd is unusually high.

You get senior bankers, consultants, investment firms, private-equity people, venture-capital firms, and technology providers all in one place.

It becomes two and a half days of great conversations with very smart people.

What were some of your highlights?

Before we get to the conference itself, I have to say that you experienced something with me for the first time: my first rock concert.

The Killers.

That was an amazing way to kick off one of my top three banking conferences.

Then we had to start with KBW’s Thomas Michaud, who always gives an incredible state-of-banking presentation.

He manages to pack an unbelievable amount of data and perspective into one session.

One point that stood out to me was his discussion of how the market rewards consistency.

That connected with something we talk about constantly at Cornerstone: pick your lane and own it.

Almost every conversation I had at the conference came back to the same idea.

Double down on your niche. Build infrastructure that actually works. Then scale it intentionally.

There was also a lot of conversation about fintech partnerships becoming more aggressive.

When you look at Capital One acquiring Brex or Fifth Third building banking-as-a-service capabilities, it reinforces the idea that if you can’t beat every competitor yourself, sometimes you partner with or acquire the capability.

Those were some of my biggest takeaways.

First, kudos to Thomas Michaud. His presentations are always impressive.

But this year, the qualitative points actually stuck with me more than the numbers.

At one point he described a “generational change in regulation.”

He walked through a timeline of 2025 that included leadership changes at agencies such as the FDIC, Treasury, OCC, and Federal Reserve, along with several regulatory and legislative developments.

His point was that the cumulative effect represented more than incremental change.

He said banks are increasingly being regulated not only by asset size, but by asset size and business model.

That is a very important distinction.

He also argued that banking innovation is becoming acceptable again, and that regulators are placing renewed emphasis on banks as engines of economic growth.

At one point he said deregulation has enabled banks to play offense.

That line really resonated with me.

For years, so much of the industry has been operating defensively around regulatory pressure.

A shift toward offense changes how executives think about growth, acquisitions, technology, and new business models.

I remember thinking, “Somebody should probably tell Elizabeth Warren,” because I’m not sure she’s going to agree with all of that.

But the broader shift in regulatory tone was one of the biggest things I took away.

Absolutely.

The changing environment is expanding what institutions are willing to discuss.

We spent all last year talking about the GENIUS Act and what it could mean for stablecoins, new charters, and payments.

Now institutions are thinking more seriously about where they can expand and where they can take calculated risks.

And hats off again to Thomas for covering so much in one session. It was standing-room only.

Another theme throughout the conference was AI.

There were several sessions on artificial intelligence and emerging technology, but I left with a very specific feeling.

There’s an old Paul Newman movie, Cool Hand Luke, with the famous line, “What we’ve got here is failure to communicate.”

That’s how I felt about some of the AI conversations.

One session started well. The moderator correctly said there are different types of AI and asked the panelists to explain the distinctions.

Kudos to her.

But the panelists didn’t answer the question clearly. They went in different directions and muddied the terminology.

Another session showed a slide defining categories of AI, and I looked at it thinking, “Some of these categories overlap, and some of these things aren’t AI technologies at all.”

That failure to communicate creates real problems.

Executives leave confused, vendors use inconsistent terminology, and institutions end up making poor decisions.

I recently got a request from a banker who wants Cornerstone’s help with AI.

Their first questions were about policy, security, controls, and protection.

Those things matter, but I think that is the wrong place to start.

Here’s the analogy I used.

Stacey, imagine you inherit a beautiful three-acre piece of land. It’s lush. You can do anything with it.

What is the first thing you do?

You figure out what you want to do with the land.

Exactly.

You don’t immediately build a giant fence and security system before deciding whether you’re building a home, a farm, a park, or something else.

But that is how some institutions are approaching AI.

They jump directly to policy and controls before defining the business objective.

There was another failure-to-communicate theme around growth.

Almost every board meeting I did last year asked me to talk about growth.

And many conversations at Acquire or Be Acquired were also about growth.

But I realized that everybody meant something different.

One banker meant zip-code analysis and finding underpenetrated markets.

Another meant new products.

Another meant acquisitions.

At a conference literally called Acquire or Be Acquired, you can guess which version came up a lot.

So the industry has a terminology problem more broadly.

We use words such as AI, growth, engagement, innovation, and data as though everybody means the same thing.

Often they don’t.

I’ll stay on that soapbox with you.

Our colleague Al Dominick made a similar point in a session about scaling with stability.

He asked two questions that probably belong in every boardroom:

What type of business are we actually running?

And how do we define growth?

Those questions sound simple, but they force strategic clarity.

The AI session we attended was underwhelming for the same reason you described.

It was standing-room only. People were taking notes because they wanted something practical to bring back to their institutions.

But if the conversation stays at “AI does this” and “AI does that,” we aren’t giving them enough.

With your land analogy, the real question is: what is the objective?

What do I want to grow? What problem am I trying to solve? What manual process could I automate? What capability am I trying to build?

That brings us to another recurring theme from the conference: data.

Steve Bohanon from Alkami made a point that the smarter bank of 2030 is going to be a data leader.

Mark DeFazio at Metropolitan Commercial Bank, who won one of our GonzoBanker Awards, made a similar point in a session moderated by our colleague Brad Smith.

Mark said that when he started banking, he never imagined he would eventually need data scientists on his team.

But that is exactly what happened.

He hired senior talent and started building a data capability inside the bank.

Steve also said that when he visits institutions, he is often surprised by how many do not have a clear data leader.

That made me corner you during the conference and ask: if you’re a bank or credit-union CEO, what should you look for in a chief data officer?

Should it be someone from lending? Operations? Technology? Analytics?

And your answer was essentially, “I don’t care. I want someone who gets things done.”

Exactly.

And this connects right back to our terminology problem.

People constantly say, “Data is important.”

What kind of data?

Customer data? Transaction data? Operational data? Market data? Competitive data? Credit data? Quantitative or qualitative data?

Be specific.

Cornerstone published two reports last year around what we called Data IQ and Data EQ, with the second focused more on execution quality.

The point is that institutions need to stop talking about “data” as one giant thing.

Poorly defined data strategies are one reason AI projects fall short.

Then there’s the organizational issue.

I’ve never been a huge fan of the “chief fill-in-the-blank officer” approach.

Sometimes organizations create a title because they don’t know how to solve the underlying problem.

JPMorgan recently announced a chief data and AI officer for its payments group, following broader executive moves around data and AI.

It reminded me of the Spider-Man meme where three Spider-Men point at one another.

I relabeled them chief data officer, chief AI officer, and chief digital officer.

If you have all three and the responsibilities overlap heavily, you may have an organizational problem.

Who owns the budget? Who owns the decisions? Who owns the outcomes?

When you kept asking what background the ideal person should have, I punted because I genuinely don’t care as much about the functional pedigree.

I care whether the person is an effective change agent.

Can they build relationships with the rest of the C-suite? Can they work across business lines? Can they get decisions made and projects implemented?

I’d rather have someone who can move the organization than someone who has spent 30 years studying data but cannot get anything adopted.

We need to get these conversations down from 150,000 feet to the curb.

And for institutions listening, the practical data stack still matters.

Data strategy. Data governance. Data warehousing or platform architecture. Those are foundational.

If you want to build AI or automation on top of weak data, you’re building on a bad foundation.

Now I know we wanted to touch on Affirm and Fiserv.

Fiserv recently announced a broader integration with Affirm to bring buy now, pay later capabilities into the Fiserv ecosystem for banks and credit unions.

The relationship between the two companies actually goes back several years, so this did not come out of nowhere.

Our colleague Tony DeSanctis reacted by asking whether banks are inviting the fox into the henhouse.

I think that is a legitimate question.

I have been telling banks and credit unions for years that buy now, pay later is becoming an ingrained product and behavior for younger consumers.

A lot of people still describe BNPL as some kind of evil credit product.

I disagree.

Buy now, pay later is basically an entry-level credit card.

Instead of giving a person a $5,000 revolving line, you approve them for one purchase. They repay it. Then they may get approved for another.

Over time, companies such as Klarna and Affirm build repayment history, underwriting knowledge, and a deeper customer relationship.

That means they could become major credit-card competitors over time.

And there are other technology providers that can help banks offer BNPL without giving the customer relationship directly to Affirm or Klarna.

I’m not criticizing Fiserv or Affirm for making the deal. If I were them, I’d probably do it too.

But financial institutions need to think carefully about who the long-term competitor is.

We recently surveyed consumers privately and asked whether they would use buy now, pay later if their primary bank or credit union offered it.

Across generations, a very high percentage said yes.

That tells me the opportunity exists for banks to own this relationship themselves.

I’m going to disagree with you a little.

I actually think the Affirm partnership can be very useful for institutions that cannot build this capability quickly themselves.

Tony has described BNPL as a better mousetrap, and I think that framing works.

It is fixed-term credit with a defined duration and clear repayment structure.

Large banks are already normalizing installments. Chase offers split-payment capabilities at scale.

Many Fiserv clients cannot replicate that quickly without a partner.

And fintechs have a history of taking familiar banking products, improving the experience, and giving them a new label.

Chime’s SpotMe is basically a modernized overdraft-protection experience.

So when banks are planning for growth and younger consumers, they have to ask what capabilities they need and whether partnership is the fastest responsible way to get there.

At the conference, Mark Zausmer from Skadden made a point that stuck with me: there is a real threat in doing nothing.

The number of banks continues to shrink.

Jack Henry’s Joshua Jordan also talked about having a playbook before entering a fintech partnership.

What do we offer today? What do competitors offer? What do we wish we had? What is actually on our roadmap?

For credit unions, maybe the answer is building capabilities through a CUSO. Maybe it’s a direct technology partnership.

But sitting still is also a strategic choice, and often not a good one.

I hear you. We’ll see how it plays out.

There’s one more topic we wanted to cover, but we’re looking at the clock and don’t want to keep people too long.

So let’s make a promise.

Next episode, we’ll talk about Capital One and Brex, even if the news is a little older by then. Let’s put that first on the agenda.

Stacey, it was great seeing you this week. It was great hanging out with you, Sam, Al, and hundreds of other people at the start of conference season.

And it was really nice doing it in warm weather.

Thanks everybody for listening.

And I don’t care what people say about you, Ron. You’re a rock star.

Stop.

If you enjoyed today’s episode of What’s Going On in Banking, follow and subscribe wherever you listen, whether that’s Spotify, Apple Podcasts, or YouTube.

Thanks for tuning in.