Meet the team: David Bholat

Could you outline your career path so far?

Before I joined Faculty, I led and launched analytical functions at Barclays and the Bank of England.

What motivated you to start a career in the financial services industry?

Motivation is a complex thing. Adam Smith, the great Scottish political economist, is often and incorrectly attributed with the view that we are motivated by self-interest. In fact, what he meant is that we are motivated to do good by our self-consciousness- we care about what others think about us and aspire to achieve recognition from them.  

In my case, I suppose I’ve always been motivated to receive recognition from as broad an audience as possible by focusing on issues that really matter to people. One of those issues is money. It concerns everybody one way or another. How do you save more? How can you earn more? Some people desire it, others detest it. Money is an object of almost universal fascination.

I started my career in financial services because I was fascinated by where money comes from. How is it produced? Through what mechanisms? On whose terms? I started my career during the financial crisis of 2007-08 and that motivated me to want to improve our monetary and financial system. It was clear then that, in the run up to the crisis, the industry had failed to serve its customers well. I got my start working in data and analytical roles. And today, whether it’s to help consumers better manage their money or navigate an increasingly risky world with more confidence, there’s so much AI can do to help financial institutions and their customers. That’s what motivates me now.

How do you see AI changing financial services in the near future?

I believe there are four key opportunity areas for financial services in the near term. 

The first relates to improving operational efficiency. One beneficial aspect of the recent inflationary environment for financial institutions has been rising profitability. However, as the macroeconomic cycle begins to turn– as base rate comes down- so will interest income, while inflation in operational costs will prove stickier. AI can help combat rising cost-to-income ratios by automating manual tasks and optimising operational routines.

The second is customer service. Consumers are today empowered by unprecedented access to information on alternatives and the ability to move their money easily through mechanisms like the Current Account Switching Service. So firms need to step up their game to retain customers. In addition there are higher regulatory expectations on firms such as the FCA’s Consumer Duty regulation. Once upon a time, regulators were solely concerned with financial services not causing any harm. But today firms need to provide evidence that they are actually driving good outcomes. In an industry where price and product strategies are easily copied, service becomes arguably the key differentiator. AI can help by unlocking new insights from data to power personalised customer service propositions. 

The third area of opportunity is navigating uncertainty and volatility. The world today is more uncertain than in the recent past, making it difficult for people and organisations to plan for the future. Domestic turmoil, international conflicts, climate change, pandemics – you name it, we’re facing it. AI can help financial institutions navigate this uncertainty through simulations, to inform business planning and personal financial planning under a range of potential future scenarios.

The fourth and final way I see AI changing financial services in the near future is bolstering the fraud and cybersecurity defences of firms. It’s often said that AI can increase productivity. That’s true, but it’s morally agnostic. It can increase the productivity of bad actors as well as the good guys and gals. We’re already seeing bad actors use GenAI to massively increase the amount of disinformation and cyberattacks on financial institutions. The good news is that we can also use AI to identify and shut down these kinds of operations.

What are the primary challenges that organisations in the financial services industry face, as they bid to innovate with AI responsibly and ethically?

I think there are two key challenges. They have less to do with AI and more to do with the humans using it. 

The first is for boards and executives to gain confidence and assurance around AI safety. AI models are complex, and organisations understandably want to ensure a measure of transparency. There are techniques that can ensure this to some extent, but not perfectly, particularly when it comes to GenAI. So firms need to accept a little bit of risk but often seem reluctant to do so. My challenge is that they should hold AI to the same standards as humans. Humans are maybe the ultimate black boxes. Neurobiologists still aren’t sure exactly how the brain works. What we do know is that our decisions are a complex function of social conventions, personal experiences, momentary whims and other factors. Yet we still trust humans to make decisions. We should feel similarly comfortable using AI to make decisions, especially if we have evidence that it drives better outcomes.

The second challenge is for organisations to not get stuck in pilot purgatory. In part, that can be avoided by working with partners that possess the right software and machine learning engineering skills to put models into production. But pilot purgatory can also be avoided by thinking more strategically. It’s easy to create a backlog of AI use cases and to get seduced into thinking progress is being made just because the list is growing. But it’s much harder to go beyond point solutions to thinking how they stack up into an enterprise programme that delivers improvements greater than the sum of the project parts. Of course, financial services firms don’t need to boil the ocean when they get started. They just need to make sure the technology they build or invest in is tightly scoped, connected to business outcomes, leverages learnings from prior precedents, and is interoperable. In this way, they can build in a way that delivers incremental value but ultimately a significant state change.

If you could have dinner with one person from the past or present who would it be?

I grew up in an extended family that included my great-grandmother, who I lived with until her death at 95 when I was 18. She used to make me dinner every night. It would be nice to have dinner again with her after so many years.