AI: The new ‘Risk Revolution’ and what that means for teams and deals of the future…
AI is moving fast into banking risk systems and the industry is right to be both excited and cautious. Here’s why the conversation needs to happen now.
There’s a quiet revolution happening inside the financial services industry, particularly the risk functions. Not yet the loud, headline-grabbing kind, but the slow, incremental kind. The kind where a decision that once took a team of experienced people several days may soon take minutes. Where judgment built over decades could be replicated…or challenged, by a process nobody in the room fully understands. Not yet. But the direction of travel is clear.
AI, of course, is not new. Machine learning has been quietly embedded in fraud detection and compliance screening for years. But what is new is its visibility and its ambition. AI has earned its stripes in the back office, but the complex, judgment-intensive corners of banking - the credit committees; the leveraged finance desks; the project finance teams; the risk functions where experience and instinct still win deals - AI has not truly arrived there yet. What is arriving is the infrastructure, the investment, and the intent. The C-suite mandates, the eight-figure technology budgets, the Chief AI Officers now sitting alongside CFOs and CROs, all of it is pointed in one direction. For the professionals working in these functions today, the honest message isn’t that AI is here. It’s that it is coming and probably sooner than the industry is ready for.
The question isn’t really ‘if’ anymore. It’s at what cost and who carries the consequences when it gets it wrong.
Faster decisions. But on whose terms?
On the credit side, the appeal is obvious. Speed, consistency, the ability to consider far more about a borrower than a traditional process ever could. At the retail and consumer end, early AI-assisted decisioning tools are already being tested. But in corporate, leveraged and project finance - where deals are complex, bespoke and high-stakes, human judgment still leads. When that changes (and the trajectory suggests it will) the question of accountability becomes critical. When a decision goes wrong, when risk declines or approves a credit submission it shouldn’t have, who explains it? Who owns it? The model doesn’t sit in front of the regulator. A person does. And if the assumptions incorporated into that model reflect the world as it was rather than the world as it is, those assumptions will play out at a scale and speed no human team ever could.
A new kind of risk, hidden inside the solution…
AI doesn’t eliminate operational risk, it transforms it. A person makes a mistake in isolation. A flawed process makes the same mistake everywhere, simultaneously, before anyone has noticed something is wrong. The risk functions most likely to feel this first are not hard to identify - credit risk, market risk, liquidity risk, operational risk, model risk and compliance are all in the crosshairs. In funds, portfolio risk monitoring and counterparty risk assessment are already seeing early AI integration. It is only a matter of time before the tooling moves from augmenting these functions to actively shaping them.
The institutions getting this right aren’t the ones moving fastest. They’re the ones asking the harder questions: Who owns this? Who challenges it? What happens when it fails? Because in banking, where trust is essentially the product, the reputational cost of a high-profile AI misstep can outweigh years of efficiency and trust gained. There are genuine wins here too, nobody should pretend otherwise. But the wins require the right people around them, asking the right questions, with the authority to act on the answers.
Moving faster isn’t the same as moving better…
For those working on more complex, structured transactions, AI is beginning to offer tools that could compress timelines significantly - research assistance, document review, comparable analysis, even early-stage modelling support. Tasks that currently consume days may soon take hours.
For talented people in these roles, that should ultimately be liberating. More time for the judgment calls that actually matter, like understanding the strategic context of a deal, building the relationships that bring mandates in the door. The things no model can do.
But there’s a risk of false confidence. A well-presented output can look authoritative even when the underlying assumptions don’t fit the situation. Speed without scrutiny isn’t an advantage…it’s a liability dressed up as progress.
It’s not just finance. It’s everyone, everywhere, all at once…!
I asked a group of friends recently across banking, marketing, advertising and sales - one simple question: are you using AI every day? Every single one said yes. Not experimenting. Actually using it. And increasingly not just for emails and presentations, but agents running whole workflows in the background.
In banking, that shift is moving steadily toward the roles that will ultimately effect roles we recruit, the investment analysts, the project finance associates, the DCM and syndications professionals and the leveraged finance teams. AI won’t replace the judgment these people carry. But it will change what’s expected of them. And when it does, that changes everything about how you hire.
This isn’t a niche technology story anymore. It’s also a workplace story and in industries like banking, where the work has traditionally been people-intensive, process-heavy and relationship-driven, the implications for roles (and the people in them) will be significant.
Some jobs will get smaller. Some will get bigger. Some will disappear entirely and new ones will emerge that don’t have a job title yet. The roles that survive and thrive will be the ones where human judgment, accountability and relationships cannot be automated away and there are plenty of those across banking and finance. But the people in those roles will look different. They’ll need to be comfortable working alongside tools that move faster than they do and confident enough to push back when those tools get it wrong.
The question isn’t whether AI will change the shape of your team. It will. The question is whether you’re being intentional about preparing for it or just waiting for it to happen to you. The talent market is changing. Less quietly and quickly. AI is raising the floor on what’s possible. But it’s also raising the bar on what’s valuable in a person.
And if you needed any further proof that this is no longer a distant conversation — some of the world’s biggest financial institutions have already appointed their first ever Chief AI Officers in the last 12 months:
These aren’t technology hires. They’re C-suite mandates. When banks start putting AI leadership at the same table as the CFO and the CRO, the question of how it reshapes the teams beneath them stops being hypothetical. That’s the conversation we are already having and if you’re not having it yet, you probably should be.
What’s your opinion?
Contact Rob Hockedy rhockedy@jmes.com.au / +61 2 9235 9470
