We Find a Way: What History Tells Us About AI in Banking
I’ve been working in fintech and digital banking for 25+ years, and I’ve heard “that will never work in banking” more times than I can count. For example: "No one will ever apply for a $1M mortgage online." Sometimes the concern was legitimate. Sometimes it was just fear of change or overfamiliarity with the way things have always been done. But here’s what I’ve noticed: when something is genuinely BETTER for the consumer, we find a way to make it work.
Not overnight. Not without friction. But we find a way.
There's been a lot of talk recently about the impact of Chase blocking AI Agents like Muse. And its understandable - there's a big question about accountability. If an AI Agent moves money incorrectly, who's responsible? The bank? The customer that set-up the connection? The software provider that created the agent?
But - with all that said, I wanted to take a different angle on this conversation. One of inevitability. This is going to happen. I think that’s the best lens for the AI conversation in financial services right now. So let’s look at some history.
Uber in New York
Let’s start with a recent example I remember well. In 2015, New York City proposed capping the growth of Uber’s fleet. Taxi interests were concerned, and city leaders had real questions about congestion and driver pay. Then riders spoke up, and the proposal was dropped.
That wasn’t the end of the story. In 2018 the city paused new licenses for ride-hail vehicles and put pay standards in place for drivers. So Uber didn’t win unrestricted access, and it didn’t get banned either. On the other side, the value of a taxi medallion in NYC went from betwen $1M and $1.3M to less than $200k. The city found a way to keep something consumers clearly wanted, with guardrails, despite protests from the powerful taxi lobby groups. It was good for the consumer, so we found a way.
Banking has done this before
Here’s the part I love. Our own industry has a long track record of pushing back on change, and then finding a way when the consumer benefit was clear.
Discount brokerage (1975). Fixed trading commissions ended on “Mayday” after the SEC stepped in, despite resistance from many established firms. Costs for retail investors came down, and decades later we got to zero-commission trading.
Index funds (1976). When Vanguard launched the first index fund for individual investors, plenty of people in the industry dismissed it. It was nicknamed “Bogle’s Folly.” Today, index funds are a cornerstone of how everyday people invest. My favourite podcast Acquired did a detailed episode on this story.
Automated underwriting (mid-1990s). When Freddie Mac and Fannie Mae introduced automated underwriting systems, there were fair questions about whether software could judge creditworthiness fairly. We didn’t abandon the idea. We built rules around it, including fair lending requirements and testing. I think this is the closest parallel to AI.
Remote deposit capture (2000s). Banks worried about duplicate deposits and fraud. Then the Check 21 Act (passed in 2003) enabled image-based check clearing, and a few years later we had mobile deposit. Now try telling a customer they have to drive to a branch to deposit a check!
Open banking. The UK’s regulator ordered the biggest banks to share customer data (with consent) through secure APIs, and Europe’s PSD2 did something similar. In the US, the debate over data access rights and who pays for what is still working itself out. But the direction of travel is clear: consumers want control of their own data, and the industry is building the plumbing to make it happen.
It’s not just banking
Other industries have grappled with change in the same way and found a path forward because, again, it was good for the consumer.
Mobile Number portability. Before 1996 (and implementation in 2003), switching mobile carriers meant giving up your phone number. Carriers had little incentive to change that. Regulation required it, and competition got better for consumers.
Telehealth. For years, reimbursement and licensing rules held it back. Then the pandemic forced the issue, and a lot of what we learned stuck around.
The pattern
Look at that list and you’ll see the same three things every time:
The initial concerns were REAL. Fraud, fairness, safety, competition. Nobody was being silly.
The consumer benefit was obvious. Lower cost, more convenience, more control, faster decisions.
We found a way. Standards, regulations, testing, and new ways of operating. Not a ban, and not a free-for-all.
So what about AI?
Don’t get me wrong. The questions about AI in financial services are legitimate. Explainability. Bias in lending decisions. Data privacy. Model risk management. Accountability. Banks and credit unions are RIGHT to take those seriously.
But I think history tells us these are questions about HOW, not IF.
Think about what AI can do for consumers: faster answers when they have a question at 10pm on a Sunday, less paperwork at account opening and loan application, better detection of fraud, and more helpful guidance on their money. That’s the stuff people want from their bank. When the benefit is that clear, the pressure to make it work only grows. Is there a risk? Of course! Banks and Credit Unions are concerned about further 'disintermediation' of their brand from the provision of finanical services.
But here’s the good news for banks and credit unions: the institutions that engage early get to help shape the guardrails, instead of having them shaped for them. And the institutions that are open to this inevitability and prepared for an agentic world (but not too prepared too early) will survive this seismic shift.
I’ve seen this movie before. About 8 years ago, my team at Avoka worked with Samsung on a prototype that let you store your identity on your phone and use your fingerprint to share it with a bank for KYC. It never went to production. Fast forward to today, and Apple’s Verify with Wallet is doing something very similar. The idea was right. The timing and the ecosystem just needed to catch up.
AI in financial services will be the same story. It will happen, because it’s good for the consumer. The real question is who will help figure out the “way.”
Comments