When AI Breaks Trust in Banking

Zachary D'Jimas

What if an algorithm prevented you from having your appendix removed? Think about that for a second. It sounds like a hyperbole, right? A worst-case scenario from some science fiction movie. But it is a very real framing for where we are heading. Let me paint three pictures for you, three entirely realistic, everyday scenarios that are happening in some form or fashion right now across our society.

Zachary D'Jimas

First, imagine a patient. They arrive at an emergency room in severe, agonizing pain. The doctor, a trained professional with fifteen years of clinical experience, stands right there, looks at the patient, and believes surgery is immediately necessary. But the insurance approval engine, a quiet piece of software running on a server hundreds of miles away, denies the procedure. Why? Because the patient's specific case doesn't meet a narrow, pre-programmed, automated threshold. The patient isn't speaking to a doctor or a nurse, or even an accountable human decision-maker. They are speaking to a queue. They are waiting on a system that has no capacity for empathy.

Zachary D'Jimas

Now, picture a family. They have banked with the exact same institution for twenty-two years. Their parents banked there. They apply for a mortgage after doing- well, after doing everything right. Stable income, good credit, a deeply responsible financial history. But their application is declined. Instantly. By an automated credit model. And when they ask why, no one- not the branch manager, not the loan officer- no one can clearly explain the decision. The family doesn't feel rejected by a bank. They feel... well, they feel erased by a system.

Zachary D'Jimas

And finally, think about an elderly customer who suddenly loses access to her funds because a fraud model flags unusual activity. In reality, she is trying to pay for urgent home care, or medication, or some family emergency. The bank's response? They say the process is automated. The call center employees are literally locked out from overriding it. This customer spent decades trusting this institution, and now, at the absolute moment that trust matters most, no human seems empowered to help her.

Zachary D'Jimas

That is the question we are facing now. Not whether AI belongs in banking, insurance, or financial services. It does. I want to be very clear about that. The real question is whether institutions that were built to serve people will still recognize people when the spreadsheet says humans are too expensive.

Zachary D'Jimas

You see, banking, insurance, and financial institutions... they aren't ordinary businesses. They aren't like software companies selling a subscription, or retail brands selling shoes. Like healthcare, they sit close to the basic, fundamental needs of human life. Shelter. Savings. Credit. Recovery from a disaster. Protection for your family. Retirement. These are the pillars of human stability. And when these institutions remove human judgment from critical decisions only to reduce cost, to improve short-term margins, or to satisfy Wall Street's pressure, they create a very silent, very dangerous kind of damage. A damage that doesn't show up immediately in quarterly results.

Zachary D'Jimas

That hidden asset they are burning through? It's trust. And let's not make the mistake of thinking trust is a soft, warm-and-fuzzy corporate metric. Trust is not soft. Trust is economic infrastructure. It is the grease in the gears of our economy. Trust lowers transaction friction. Trust keeps customers loyal when a competitor lowers their rates. Trust protects your brand during a systemic crisis. And trust is what allows an institution to recover from its inevitable mistakes. But trust is built slowly, over decades, and it can disappear in a single automated afternoon.

Zachary D'Jimas

I was recently reading a white paper published by Harvard Business Review Analytic Services, titled 'The Human Factor: AI-Powered Customer-First Strategies in Banking and Financial Services.' It is a fascinating piece of work, and it really got me thinking about this delicate balance. The paper highlights how artificial intelligence can genuinely optimize personalization, speed up fraud detection, support financial wellness, and elevate customer service. But the authors make a vital point: long-term success still depends entirely on trust, on human judgment, transparency, and customer-first leadership.

Zachary D'Jimas

When you look at the actual data, the market leaders who will thrive in this AI era aren't simply the ones who deploy the fastest algorithms. They are the ones who understand how to use technology to strengthen, rather than replace, the human relationship. Think about it. When a bank, insurer, or financial institution removes the human from the decision, it may also remove the person from the relationship. If your interaction with a brand is completely indistinguishable from an interaction with a basic utility grid, why should you stay loyal? What is keeping you there?

Zachary D'Jimas

There is a massive, systemic difference between using AI to augment your people and using AI to shield your institution from accountability. AI should help professionals see customers more clearly. It should not become a curtain that hides the customer from the institution. If an advisor uses a generative model to synthesize complex data so they can have a deeper, more meaningful conversation with a client, that is augmentation. That is beautiful. But if a risk model is used to auto-deny an exception without any human recourse, just so executive leadership can boast about headcount reduction, that is not innovation. That is a strategic cop-out.

Zachary D'Jimas

What we are seeing in the market right now is the accumulation of what I call 'trust debt.' Just like technical debt in software development, trust debt occurs when you take shortcuts today that you have to pay back with interest tomorrow. Wall Street can measure headcount reduction. It has a much harder time measuring the value of a customer who still believes you will be there when life becomes difficult. When you cut the staff who handle exceptions, you might save a few million dollars this quarter. But what happens when your customers realize that your system doesn't care about their individual, complex lives?

Zachary D'Jimas

Efficiency is valuable. But efficiency without judgment can become institutional abandonment. It is equivalent to a hospital cutting essential patient safeguards, like nurse-to-patient ratios or manual checklist verifications, simply to meet some abstract balance-sheet target. Sure, the numbers look great on paper for a brief moment. But the underlying system becomes fragile, cold, and prone to catastrophic failure. An algorithm can calculate probability. It cannot understand dignity. And the most dangerous failure in financial services is not always a system outage. Sometimes it is a trust outage.

Zachary D'Jimas

I want to give you, the listener, some practical tools here. If you are trying to figure out if your own bank, or your insurer, or your investment firm is crossing this line, here are some warning signs. Let's call this: Signs Your Trusted Institution May Be Choosing Automation Over Care.

Zachary D'Jimas

First, you can no longer reach a human for complex or urgent issues. The system forces you into an endless loop of automated chat menus. Second, decisions- like a loan denial or a premium increase- are made incredibly quickly, but explained incredibly poorly, relying on vague template language. Third, appeals or disputes feel like just another automated workflow where you receive the same robotic response. Fourth, when you do finally reach a human employee, they sound completely powerless to help you, often saying, 'The system won't let me change that.'

Zachary D'Jimas

Fifth, policies are applied rigidly, even when your situation is clearly unusual and warrants a simple, common-sense exception. Sixth, the institution talks constantly in its public marketing about efficiency and technology, but rarely about service quality or human care. Seventh, they assume your customer loyalty is guaranteed, treating your history with them as just another data point. And eighth, human exception handling is treated by their management as a costly problem to be eliminated, rather than a vital trust-building function.

Zachary D'Jimas

So, how do we fix this? If you are a leader inside one of these organizations, how do you steer the ship back toward human-centered trust? Strong, ethical leadership requires very specific actions. First, you must use AI to support your employees, not to erase their accountability. Give your front-line staff the power, and the budget, to override algorithmic decisions when common sense demands it. Keep humans in the loop for any decision that has a high consequence on a customer's life, shelter, or health.

Zachary D'Jimas

Second, you need to measure what actually matters. Don't just track cost savings and processing speed. Measure trust. Measure complaint quality, exception resolution times, and- most importantly- customer harm. Protect your institutional knowledge. When you lay off seasoned workers in the name of automation, you lose the very people who understand the unwritten rules of customer relationships. Build clear escalation paths with real human authority, and treat customer dignity as a core strategic asset, not a luxury.

Zachary D'Jimas

At the end of the day, we have to remember the fundamental relationship between human design and technology. AI is an extraordinary engine. It can process vast amounts of data, find patterns we would never see, and accelerate our workflows. But humanity must always remain the steering wheel. An engine without a steering wheel is just a fast way to crash.

Zachary D'Jimas

The future of financial services should not be humans versus AI. It should be humans with AI, serving other humans better. But if institutions forget why they were trusted in the first place, they may discover something painful: the customer relationship was never stored in the model. It was stored in the belief that when life became complicated, someone would still care enough to listen.

Zachary D'Jimas

If this episode made you think differently about trust, banking, insurance, or the future of human judgment, subscribe to The Human Workforce. This is where we examine the future of work, technology, leadership, and the human decisions that still matter most. I'm Zachary D'jimas. Thank you for listening, and let's keep building a workforce that works for all of us.