Carey Ransom is a SaaS entrepreneur, executive, investor and advisor. He is president of Operate and managing director of BankTech Ventures, a strategic investment fund backed by community-bank leaders and focused on technologies that strengthen their ecosystem.

Strange as it sounds today, the arrival of ATMs in my childhood sent bank executives into full panic mode. Some thought they’d finally replace tellers. Others were convinced customers would reject the machines entirely and even paid out small bonuses to encourage early use.

Both were wrong.

What actually happened? The cost of operating a branch dropped so much that banks opened MORE branches. Tellers didn't disappear. Their jobs transformed. They stopped counting cash all day and started solving problems, selling products and building deeper relationships.

I think we're at a similar moment with AI, but the transformation will cut much deeper. And here's why community banks might actually have a hidden advantage that nobody's talking about yet.

AI cuts straight through the dense layers big banks rely on, from committees to endless reviewers. Community banks aren’t trapped in that architecture, and that gives them a rare chance to use AI for faster, closer-to-the-customer decisions that many big banks can’t match.

The blind spot in today’s AI conversations

Every bank I talk to asks the same question: "How do we use AI to do what we're already doing, but cheaper and faster?"

That's the easy but wrong question. It's like asking in 1995: "How do we use the Internet to make our fax machines work better?"

The right question is: What becomes possible when every frontline person in your bank has capabilities they've never had before?

Imagine your loan officer sitting with a local contractor who wants to expand his business. Right now she probably takes the application, sends it to credit analysis, waits for their report, sends it to the loan committee, waits for their meeting and comes back to the contractor a couple weeks later.

Now imagine she has AI that can do sophisticated financial modeling, scenario analysis, industry research, and risk assessment while she's still sitting with the customer. It's not replacing her judgment but augmenting it. She can explore "what if" scenarios together with the contractor, understanding the business in ways that would have required an entire team before.

Here's the uncomfortable question: If she can do all that, what's the credit department for? What's the loan committee for? Hmm.

The fork in the road

This is where banks are going to split into two very different futures.

Some banks will see AI's power and think: "This is dangerous. We need MORE controls, MORE oversight, MORE approval layers." They'll create Chief AI Officer positions, form AI governance committees, build approval workflows for AI recommendations, hire consultants to validate their models and they'll get distracted from the huge potential, learning and growth opportunities.

They'll probably get all the AI costs, all the complexity, and very little of the benefit. Their talented people might even leave because they'll be drowning in more process. They get slower, not faster.

Other banks will recognize something profound: If AI can handle information processing and routine analysis, then the entire reason most of their hierarchy exists just evaporated.

Let's look closer. Why do banks have layers? Partially it's about risk management, sure. But a huge part is information processing. Moving data up the chain so someone with enough context can make a decision, then moving that decision back down.

How should they organize when that's not necessary anymore?

What actually changes

Here's where it gets interesting, and why I think community banks might have an unexpected advantage.

The big banks are stuck. They've got massive compliance departments, risk management frameworks, layers upon layers built up over decades. Their org charts look like architectural diagrams. Unwinding that is like trying to renovate a skyscraper while people are living in it.

Community banks? You're smaller. More nimble. And you've got something the big banks desperately want but can never really have: you actually know your customers and your communities.

So imagine this future: Your loan officers become something more like "financial architects" for their customers. They've got AI handling the analytical heavy lifting, which frees them up to really understand what their customers are trying to build. They can be more creative problem solvers and make sophisticated decisions quickly because they have both the tools and the authority.

Your middle management layer, the people who mostly review and approve things today, will see their jobs change pretty dramatically. Some may become coaches, helping loan officers develop better judgment. Some can become relationship managers handling your most complex customers and opportunities. And some can become the people who watch for patterns that even AI might miss and improve it over time.

And here's the other part: You'll actually need fewer of them. Not because you're trying to cut costs, but because the work of coordination and oversight will largely happen through AI systems.

But you do need something new in your bank. You’ll need people whose entire job is making sure your AI systems are working correctly. They won’t review every decision the AI helps with, which defeats the purpose, but will monitor the system itself. Think of it like this: When you started using online banking, you didn't hire people to watch every transaction. You hired people to make sure the system was secure and working properly.

This is where the hierarchy question gets interesting. You don't need more vertical layers; you need a parallel function, small but critical, that watches the system’s health. You’ll want people who understand both banking and AI and can spot when models are drifting or developing blind spots. I’ve already met a few of them, and they are becoming a real superpower for their banks.

The community banking advantage

Here's why I'm actually optimistic about community banks.

The big banks are going to use AI to become even more algorithmic and more standardized. They'll optimize everything. And they'll be good at it.

But you know what they can't do? They can't know that the contractor who looks risky on paper is actually the guy who's been reliably building half the town for twenty years. They can't know that the retail shop asking for a loan is run by the woman who turned around three other businesses. They can't feel the pulse of a community.

AI done right should make you MORE human, not less.

Every minute your people spend on data entry, routine analysis, paperwork and report generation is time they’re not spending understanding their customers and their community.

Here's the thing: You can't compete with JPMorgan on efficiency. You can't compete with Bank of America on technology. But they can't compete with you on being truly local.

So use AI to eliminate everything that isn't relationship and judgment. Let it handle the plumbing so your people can be fully present with customers.

What this means

Start with one question: What are your best people spending time on that a computer should be doing? Not could be doing, but should be doing because it’s routine, analyzable and doesn’t require human intuition. And who in your bank can help them identify those tasks and start making changes?

Maybe it's credit memos. Maybe it's regulatory reports. Maybe it's data gathering for loan committees. Find those tasks and automate them aggressively.

Then ask the harder question: If those tasks disappeared, what would you want those people doing instead? More time with customers? Deeper community involvement? Developing new business relationships? How might you give people more autonomy and agency to be better value-add relationship builders?

The banks that figure out how to use AI to make their people MORE capable, MORE empowered, MORE focused on what matters are going to thrive. They'll be faster, better, and more profitable. And they'll have different structures and hierarchies than what exists in banks today.

Enjoying your read? Sign up here

The banks that use AI only to make their existing bureaucracy a little more efficient are going to fall behind. They’ll get outpaced by competitors that actually transform how work gets done.

The hierarchy question sorts itself out once you're clear on the purpose question. And the purpose is becoming more human by letting technology handle the mechanical parts.