Smart Lending, Wrong Beach: Where AI Sends Bankers to Sink or Swim

by Dain Ehring November 2025

AI is new and everywhere, it creeps into every corner of modern finance, and the more we pretend otherwise, the more it lurks, ready to bite—not unlike the monster under your childhood bed, except this one actually watches your stock portfolio. It frightens us a lot, and not in the fleeting way an ugly quarterly report does, but with the bone-deep anxiety that only existential threats evoke. It’s rarely subtle: the news headlines scream that China will leave us in the dust if we hesitate, that our jobs are poised for replacement by silicon, and that the video of your CEO on Instagram might just be a deepfake confession about next quarter’s numbers. Every email urging us to “verify our account” could just as easily be written by a bot, and sometimes, late at night, we stare at our inbox and wonder if we’ll even be employable next year.

Of course, not every fear is totally rational. Most of us work in job categories invented after our baby photos were taken, and yes, the Big Brother commercial aired four decades ago, but that doesn’t make this feel any less new and ominous. But in financial services, especially, there’s a case to be made for full-blown panic. Here, information and privacy are the currency; risk is reviewed in painstaking detail, fraud is everywhere, and regulatory mistakes aren’t just embarrassing—they can lead to orange jumpsuits and a lot of uncomfortable phone calls.

The most sobering fear? That we will adapt heroically—transforming, modernizing, rebranding—and die anyway, because the market itself moved. Imagine Sears pouring billions into revamping its catalog printing with the best technology on the planet, only to be rendered obsolete overnight by a nascent Amazon and a few million modem-wielding visionaries. In the same way, digital retail didn’t just tweak the rules: it took the ball, the field, and the referee. When I started in housing finance in 2000, we’d already spent $25 billion “modernizing,” but the mortgage playbook was untouched, still paper and folders and beige cubicles, chugging along obliviously as the world outside digitized.

Then there’s real estate—the ground zero for disruption. Why can’t a homeowner turn their property into a tokenized asset and sell it globally by breakfast? What’s really stopping a homebuyer from using a chatbot to wade through NFT listings until they find the perfect “forever home,” less the hanging sign with a realtor's photo in the front yard and the traditional banking fees? Why does regulation still look as dusty as an old courthouse when the tech world is raising tens of millions—in the same two years it takes a financial enterprise to finish arguing about software upgrades? What’s the shelf-life of our high-minded investment strategy when a tiny competitor can leapfrog with a better, blockchain-based business model tomorrow.

And as sleep disorders among executives increase, let’s discuss the AI hype machine. Every vendor promotes the One System to Rule Them All, every coder in a suburban garage now claims they will disrupt the industry by Monday, and meanwhile, enterprises are so exhausted from evaluating new products they barely have the energy to pretend they’re interested. Because honestly, it only takes a high school diploma and some free tools to generate code these days. So how do we know which AI-generated decisions will actually be helpful, and which will call a customer “probably not pregnant”—or worse, ruin our portfolio entirely? How do we make billion-dollar bets when the field’s “experts” are so young they look lost without a TikTok filter, and there’s no precedent to follow except our collective panic faces? It’s like being a deer in the headlights, only the car is a self-driving one, and you’re not sure who programmed the algorithm to drive it or if they even have a license as you sit in the back seat.

Of course, insomnia sets in not just because of business risks or technological confusion, but the real, sharp-edged fear of legal action. While federal watchdogs might lag, state agencies will gladly swoop in, using AI against you to scan for the tiniest irregularities. Suppose you get called out, good luck explaining that you trusted your chatbot more than your own compliance team. Reputation risk snowballs, customers get jittery, and suddenly you’re re-enacting the latest finance scandal for a crowd of hungry journalists. The real kicker? Even your own lawyers may shrug and blame, sighing that it was, after all, the chatbot’s fault. If that doesn’t keep you up at night, you’ve either automated away your stress or you’re simply not paying attention.

This is the new reality—messy, fast, bristling with peril, and for all its technical swagger, deeply human in the way it rattles nerves from the cubicle to the boardroom. Maybe a little panic is warranted after all. If nothing else, it keeps us from sleepwalking over the cliff. Now, if only the bots would make us a nice cup of chamomile tea and tuck us in, we could get back to pretending we’re not afraid.