AIR's Jo Ann Barefoot on AI, automation and the future of financial regulation

AIR's Jo Ann Barefoot on AI, automation and the future of financial regulation

Joe Kornik: Welcome to the VISION by Protiviti interview. I’m Joe Kornik, Editor-in-Chief of VISION by Protiviti, a global content resource examining big themes that will impact the C-Suite and executive boardrooms worldwide. Today, we’re exploring the future of government and I’m thrilled to be joined by Jo Ann Barefoot, CEO and co-founder of the Alliance for Innovative Regulation, or AIR, a nonprofit organization working globally to promote a more fair, inclusive, and resilient financial system by helping adapt financial regulation for the digital age. She hosts the global podcast, Barefoot Innovation, and is a Senior Fellow at the Harvard Kennedy School Center for Business and Government. She serves on multiple global boards and advisory bodies and is a renowned public speaker and author, having published more than 200 articles on innovation and regulation. Jo Ann, thank you so much for joining me today.

Jo Ann Barefoot: I’m delighted to be here. Thank you.

Kornik: So, Jo Ann, it’s been about five years since you launched AIR, the Alliance for Innovative Regulation, as I mentioned in the intro. Could you tell us a little bit about AIR and then how you would assess the first five years and maybe what do you see are some of the biggest challenges and opportunities of the next five?

Barefoot: Yes, I’m happy to do so. Yes, we have just passed our fifth anniversary at AIR. We are a nonprofit. We are based in the United States but active around the world and really, the idea for it came out of the bat. I’m a former bank regulator myself, begin immersing in how technology is changing the world, including finance, and reached sort of an epiphany that it’s going to be hard to regulate change at this speed and this wall breaking nature unless we really step up the focus on the regulatory process itself. We work to help regulators, financial regulators, both use technology themselves and understand the technology they’re overseeing in the marketplace. The challenges going forward are that everything is speeding up and becoming even more wall breaking and we think it’s like race against time to keep pace with the change.

Kornik: Right. I love the name because innovation and regulation are just two words that you don’t often hear together. They are often sort of opposites.

Barefoot: Yes.

Kornik: I’ll start up by asking you about your views on the future of regulation, just particularly in the financial sector.

Barefoot: If we look at the financial sector, it is changing as much as anything else around us in terms of technology, digitization, tokenization, all the trends that we’re living with, and it really matters. It’s an area where if it doesn’t work well, people get hurt and therefore it’s pervasively regulated maybe more than any other sector. The big changes, I think, and you won’t be surprised to hear me say this, are going to be dominated by artificial intelligence. The advent of generative AI, a year and a half ago, is like nothing we’ve ever seen before in terms of hitting everyone’s desk at the same time in just a few months and making everyone rethink what they are doing. So, the big trends are going to be really keeping up with that change, deploying it, and also properly regulating it so that we get the benefits of new ways of doing things but we deal with the risks, which are very, very real.

Kornik: Right. Since you brought up generative AI, let’s stay on that topic for a few more minutes. I’m curious about it’s ultimate impact. I know there was a recent AIR white paper that asked if AI is transforming the future or triggering fear or perhaps a little bit of both. How concerned are you about that generative AI future and how do we make sure it’s actually a force for good and not the other way around?

Barefoot: Yes. We definitely think it’s a double-edged sword, very dangerous for sure. There’s risks that AI and generative AI specifically is going to introduce bias into the system. We have the challenges of so called explainability, why is the AI does doing what it’s doing or recommending what it’s suggesting. We’ve got problems with data quality. We’ve got problems with model risk management. We’ve got problems with privacy. We’ve got problems with AI’s hallucinating. I mean, we don’t minimize the risks of it all, but having said that, we think the upside opportunity is huge and we have to harness it. If we put that in the framework of thinking about financial regulation, when you have a problem that at its heart is about the difficulty of not getting enough information and analyzing it or keeping up with the dynamic environment, those are the situations that AI is perfect for solving. Particularly generative AI, the ability to search large language models. If we have the time, I could give you so many examples where the way we do things today is hampered because we don’t have enough information in hand; it’s in analog form. People are doing manual data entry, I mean, regulators are doing it, risk managers and compliance people are doing it. Everyone is dealing with fragments of information in a system that’s huge and changing at a breathtaking rate and we need to harness these tools to understand what’s going on in them and get hold of that.

Kornik: Right. Are the tools in place? Do the regulators have the tools right now to protect the system and ultimately make it more resilient, inclusive, and fair? How do you balance the need for those guardrails without stifling innovation?

Barefoot: Yes. The regulators, with a few exceptions, do not have the tools now. One of the things we really advocate for, and it might sound nerdy and not interesting but it’s essential, is that the regulators have to overhaul their own technology. Not just adopt some different tools but go back and redesign their information architecture and their talent lineup. They’ve got great people who understand risks in the system but those people, the colleagues who understand technology, they need software engineers and designers. So, attracting and retaining talent, working on culture change. We are putting out a white paper on this shortly by my colleague Nick Cook, who used to head Innovation at the Financial Conduct Authority in the UK which we’re calling Regulators’ Odyssey. They need completely different tools they need to invest, and they all need to, for one thing, migrate to the cloud, which will take a little while but really has to be done.

Kornik: Right. That’s interesting also and I keep an eye out for that. So, I was listening to your Barefoot Innovation podcast, which is excellent by the way. Recently you had Elizabeth Kelly, Special Assistant to the President for Economic Policy at the White House National Economic Council, and I know she was key in drafting the Executive Order on AI by President Biden. So, I’ll ask you, do you think that the U.S. is doing enough? Do we have the right resources and focus pointed at the problem?

Barefoot: The U.S. is doing a great deal and the Executive Order led to learning and innovation really throughout the government. We are not as far along as some other places in thinking through policy toward AI and some would say that was a bad thing, some would say it’s a good thing. We’re going to need to learn to create more dynamic policy change mechanisms because whatever is adopted today on AI is going to be out of date before long because the tech itself is changing so rapidly and so are the use cases. But I would like to see the U.S. really think through a consistent set of principles for regulation of AI and be ready to put that out. I’ll mention another thing that we think is really important. At AIR, we look a lot at consumer finance, consumer protection, and financial inclusion. We think that one of the things that deserves a thought from policymakers and from the private sector is the potential for consumers, financial consumers, to have their own agents for helping them manage their financial lives. Call it a robot, call it an agent. Gartner has called this phenomenon “Custobots,” the AI devices that are going to be able to make decisions. Financial consumers don’t have enough information either and they don’t have, on average, the knowledge to fully understand the financial products that they’re looking at. If we can encourage development of AI assistants for the financial consumer who can help her pay her bills, do her budget, evaluate financial products, see through offers that may have hidden fees or terms that would be adverse to her, manage to meet her own preferences and priorities as she’s saving and trying to build wealth. If we had these and they were affordable for everyone and that was a business model that could support them, and if they were regulated to have to have a duty of care to the individual or a fiduciary duty with no secondary agenda, we could transform consumer financial markets and really reward the good actors who were trying to provide the best products and services and prices and wring out the bad actors who are not. So, we think policymakers should be thinking about that.

Kornik: Right. Very interesting and a lot of this emerging tech AI movement is being led by the private sector of course, and it seems like perhaps an area that’s right for public-private partnership. Do you see opportunities for collaboration between government entities and the private sector?

Barefoot: I couldn’t agree more, Joe. We are going to have to have private and public collaborative efforts here. I think we may need some new models for those but all these area standards, the best practices, interoperability of different kinds of systems. One example is, digital identity is emerging as the key issue in fraud, which is skyrocketing, partly thanks to AI and financial inclusion and privacy. You can’t design a systemic approach to something like that unless you have the public and private sectors working closely together, which people are trying to do but we have a long way to go to do better.

Kornik: Right. Well, Jo Ann, you’ve been very generous with your time. I just have one more question for you and that’s to take me out a few years, maybe to the end of the decade, and talk to me a little bit about what you see as the future of financial regulation in, let’s say, 2030?

Barefoot: I think the future of financial regulation is like the future of everything else. Frankly, it’s data and AI and I think by 2030, we’re going to have a lot of situations where, say, a bank examiner or a securities regulator is sitting at their desk and that he too have an AI agent that enables them to get all the information that they could possibly want just by querying, just by talking to an AI that will then bring you the answers that you have, cut data in different ways, pick up the trends, the early warning systems. I think these systems are going to remove the need to do so much regulatory work with lagging indicators at risk, which is what we have mostly today, and the whole industry is going to be able to get ahead of problems. It’s not going to be perfect, I’m not naïve, and it will be so much better today if when you speak in to see risk emerging or non-compliance emerging, you can flag it immediately and fix it immediately instead of letting it accumulate over the years. It's going to be much more beneficial to the customer and the public and also to the regulated firms. They’re going to have an easier time running their businesses in a fortuitous way.

Kornik: Jo Ann, thank you so much for your time today. I really appreciate it and I really enjoyed our conversation.

Barefoot: I do too.

Kornik: And thank you for watching the VISION by Protiviti interview. For Jo Ann Barefoot, I’m Joe Kornik. We’ll see you next time.