This is a summary of an episode of Main Street AI, an educational podcast on AI led by our founder. Join 3,700+ business leaders and AI enthusiasts and be the first to know when new episodes go live. Subscribe to our newsletter here.

TL;DR:

• Community banks are not credit unions. They're for-profit, locally owned institutions that pay taxes and serve anyone in their geography, not member-owned cooperatives.
• Carey Ransom, Managing Director of BankTech Ventures, backed by ICBA and The Venture Center, invests in fintechs on behalf of over 125 community banks who don't have time to vet a thousand startups a year themselves.
• His framework for AI adoption is "experiment broadly, build narrowly": try everything with no long contracts, then invest in building only the few things that are genuinely unique to your bank.
• His more radical advice: assume AI can do the job, and make your team prove why it shouldn't, rather than assuming a human should do it by default.
• The real driver behind bank AI adoption right now isn't hype, it's an aging workforce retiring out of roles nobody has redesigned yet.

Before we dive into the key takeaways from this episode, be sure to catch the full episode here:

What Actually Makes a Community Bank Different

Carey Ransom has a personal stake in this distinction. His family has owned shares in an Indiana community bank for over a hundred years, and that experience shaped how he describes the category now.

"I call community banks the original DeFi, because it's decentralized banking that in many cases were formed because a group of people in a local geography realized if we want to all prosper here, we have to figure out how to multiply that together."

The distinction from credit unions matters for anyone deciding where to bank or invest. Community banks are for-profit, taxpaying institutions open to anyone, while credit unions are member-owned cooperatives that require membership eligibility. Both exist to serve a specific community or group underserved by megabanks, but the ownership and tax structure differ completely. Ransom's concern is what happens if community banks lose ground: "It's not obvious that those voids will be filled," particularly when larger banks raise minimum account sizes and cut ties with longtime customers who never did anything wrong.

Why BankTech Ventures Exists

BankTech Ventures is not a typical venture fund. It's backed by ICBA (Independent Community Bankers of America) and The Venture Center, and it exists to serve the roughly 125 community banks currently invested in it, plus hundreds more it maintains relationships with.

"We've built what I would characterize as the R&D arm of the banking industry. They don't have the time or the resources to go meet and filter through a thousand different entrepreneurs a year, so they count on us to do that."

The fund has met roughly 5,000 companies and made 28 investments since it started, a narrow funnel by design. Ransom is explicit that BankTech Ventures will recommend a fintech to its bank network even when the deal doesn't make sense as an investment, which isn't standard VC behavior.

Experiment Broadly, Build Narrowly

This is Ransom's core framework for how community banks should approach AI, and it's built to run two tracks at once.

"Experiment broadly, build narrowly. You need to be experimenting, learning by doing. Don't sign long contracts. And at the same time, you might realize there are some things you want to build and own yourselves, agents that could be unique IP of your bank."

On the vendor side, he pushes fintech founders toward the same philosophy in reverse: land with something small that proves value fast, then earn the right to expand, rather than pitching a long, considered enterprise sale that community banks no longer have the patience for.

Assume AI Can Do Everything, Prove Why It Shouldn't

Ransom's most direct advice inverts the default posture most banks take toward AI.

"I think you should start with the premise that AI can maybe do everything in here, and you should prove to yourself why it shouldn't do it. Because if others push it further and are successful, they'll be wildly more efficient and scalable than you, and you've lost your competitiveness because you didn't challenge yourself as much."

He's careful to draw a line: humans stay in the loop for oversight, explainability, and accountability. The point isn't removing people, it's not assuming a task belongs to a human just because that's who's always done it.

Why the Real Deadline Is an Aging Workforce, Not AI Hype

The urgency Ransom sees in community banks isn't coming from competitive pressure alone. It's coming from retirement.

"We have folks that are going to retire a year, two, three years out, and the job they've done effectively for 10, 20, 30 years probably needs to change. They may not be the one to redefine how that should be done. That may be where AI plus newer, younger folks with a new way to think about it needs to come in."

That framing extends to customers too. Younger, mobile-first customers expect features like earned wage access or in-app investing, the kind of experience fintechs already offer, and community banks that don't adapt lose that segment by default, not by choice.

How This Works in Practice

"You are actually a beneficiary of the fact that you haven't made these huge investments in data and development infrastructure, because the cost of these solutions is far less today than it has historically been. You can be a beneficiary of that cost deflation and catch up in a way you probably never imagined."
— Carey Ransom, BankTech Ventures

Multimodal builds purpose-built AI agents for community banks and credit unions, and the build-versus-buy decision Ransom describes is one institutions are actively working through right now. If your institution is weighing which parts of an AI strategy to build in-house versus buy from a vendor, the buy-versus-build breakdown is a practical starting point.

Want more on financial services and AI? Check other episodes here.

Frequently Asked Questions

1. What is the difference between a community bank and a credit union?
Community banks are for-profit, taxpaying institutions open to any customer in their geography, while credit unions are not-for-profit, member-owned cooperatives that require membership eligibility. Both serve local or specific communities, but their ownership, tax status, and customer eligibility differ.

2. Should community banks build or buy their AI tools?
Most community banks should buy broadly and build narrowly. Carey Ransom's framework is to experiment with many AI tools without long contracts, then invest in building only the specific capabilities that represent unique IP for that bank.

3. What does BankTech Ventures do?
BankTech Ventures is a strategic investment fund, backed by ICBA and The Venture Center, that vets fintech startups on behalf of over 125 community bank investors. It functions as a shared R&D team, sourcing and filtering technology so individual banks don't have to.

4. Is AI going to replace community bank jobs?
Not by default, but the jobs will change. Ransom argues the honest starting assumption should be that AI can likely do a task, with the burden on the bank to prove why a human still should, rather than assuming automation only applies where it's obviously needed.

5. Why are community banks under pressure to adopt AI right now?
The biggest driver is workforce retirement, not competitive hype. Many banks have employees retiring within one to three years whose roles were never redesigned, creating an opening where AI and newer staff can rethink how the work gets done.

6. How does BankTech Ventures evaluate which fintechs to recommend to banks?
It screens for founders who understand banking's real constraints, sales cycles, budgets, compliance, and customer lifetime value, rather than founders chasing a venture-scale outcome that may not fit how community banks actually buy technology.