We used to deal with risks in silos, you'd have your geopolitical risk over here, your market risk over there, your cyber threats somewhere else. But that world is gone. Today, all of these forces are colliding at the same time, and they're feeding off each other in ways that our traditional models just weren't built to capture.

The organizations that thrive over the next decade won't necessarily be the fastest-growing ones. They'll be the ones that can adapt quickly while keeping their balance sheets strong enough to withstand the unexpected.

Resilience, not speed, is becoming the new competitive advantage.

I'm particularly focused on what I call the "higher-for-longer" hangover. After more than a decade of near-zero interest rates, an entire generation of risk professionals has never managed a sustained tightening cycle.

We're now watching the lagged effects of monetary policy work their way through commercial real estate, middle-market corporate debt, and private credit and frankly, we're only seeing the tip of the iceberg.

At the same time, geopolitics has become a permanent fixture in our daily stress tests. Supply chains are being redrawn along political lines, and that's not just a procurement headache; it's a concentration risk that can impair asset values within hours of a single tweet or policy announcement.

And then there's AI. I don't think we fully appreciate how AI is going to change market dynamics, not just in terms of efficiency, but in terms of correlation.

When multiple firms are using similar algorithms trained on similar historical data, they're going to react to volatility in the exact same way, selling the same assets at the same time. That's how flash crashes happen.

So the trend I'm watching isn't any single risk; it's the interconnections between them.

This is such a crucial question, because in my experience, organizations almost never fail because they missed a major, headline-grabbing risk.

They fail because they ignored a series of small, nagging warning signs that gradually became part of the background noise. It's the boiled-frog syndrome, things get hotter and hotter, but slowly enough that nobody jumps out of the pot.

The warning signs are usually hiding in plain sight. One of the biggest is the erosion of constructive dissent. When I walk into a room and everyone nods in agreement, that terrifies me more than a heated debate. Silence is not a sign of alignment; it's often a sign that people have stopped feeling safe enough to speak up. Junior staff stop escalating concerns because they assume leadership must already know, or they worry about being labeled as obstacles. That's when the first line of defense collapses.

Another signal I watch closely is the frequency of exceptions. Every organization has policies and limits, and every now and then, you need to grant a waiver for a good strategic reason. That's fine. What's not fine is when those exceptions start becoming routine, when bypassing the control framework becomes the default way of getting things done. That tells me the governance structure is being slowly hollowed out from the inside.

And perhaps the most dangerous warning sign of all is overconfidence during good times. When the portfolio is performing well, there's a natural temptation to relax covenants, increase leverage, and assume that the models have finally figured out how to predict the future. That's when risk accumulates beneath the surface, invisible to everyone until it's too late. The phrase that keeps me up at night isn't some dramatic warning; it's the quiet, mundane statement: "We've always done it this way." That phrase is the death knell of proactive risk management.

This is a question I get asked constantly, and my answer always starts with a gentle but firm reminder: start with the problem, not the technology.

I've seen too many organizations get seduced by the shiny new toy. They hear about AI, they get excited, and they start looking for a problem to fit the solution. That's backwards. You should always begin by asking yourself: What meaningful business challenge are we actually trying to solve?

Once you've identified that, the next question is whether AI is genuinely the right tool for the job. There are plenty of problems that are better solved with better processes, better data hygiene, or just a simpler piece of software. AI is incredibly powerful for probabilistic, pattern-recognition challenges like predicting customer churn or detecting fraud, but if you're trying to automate a straightforward reconciliation, you don't need a large language model; you need a good Excel macro and maybe some basic scripting.

The other thing I always emphasize is data maturity. AI is garbage-in, garbage-out on steroids. If your underlying data is siloed, inconsistent, or biased, theAI isn't going to magically fix that, it's going to amplify those flaws and present them with an air of mathematical authority that makes them even more dangerous. I often advise organizations to invest eighty percent of their AI budget in cleaning up their data infrastructure and only twenty percent in the model itself. That might sound counterintuitive, but it's the only way to build something that's actually reliable.

I think many organizations are underestimating the risks created by speed.

Whether it’s AI adoption, rapid expansion, or new product growth, the pattern is often the same: decision-making starts moving faster than governance, controls, and human judgment can keep pace.

In AI, one of the biggest risks is overreliance. Teams can become too comfortable accepting AI-generated outputs without enough challenge or independent verification. Over time, that weakens critical thinking and creates what I would call cognitive atrophy. People stop developing the instincts that come from working through complexity themselves, and when the system gets something wrong, they may not recognize it quickly enough.

There’s also the systemic risk of correlation. If multiple firms are using similar AI systems trained on similar data, they can end up reaching the same conclusions and reacting to volatility in the same way at the same time. That creates herd behavior and can intensify market stress in ways that are difficult to predict.

Outside of AI, growth creates its own hidden risks.Companies often expand into new products, geographies, or technologies before their internal systems, liquidity planning, compliance structures, or talent base are ready to support that complexity. The opportunity looks attractive on the surface, but the operational foundation underneath may not be strong enough.

What makes this dangerous is that these risks usually build quietly. They rarely appear as one major event. They accumulate through small compromises, repeated exceptions, and growing confidence during good times. By the time they become visible, they are often much harder to reverse. The strongest organizations are the ones that grow and innovate without letting speed outrun discipline.

Conversations with the executives, authors, and operators who run complex organizations.

Explore our current leadership spotlight.