C. Peter McColough Series on International Economics With Michael S. Barr

Vice Chair for Supervision Michael S. Barr discusses the potential impacts of AI on financial stability and the regulatory considerations surrounding it.

The C. Peter McColough Series on International Economics brings the world’s foremost economic policymakers and scholars to address members on current topics in international economics. This meeting series is presented by the Maurice R. Greenberg Center for Geoeconomic Studies.

SORKIN: Good afternoon, everybody. Want to welcome everybody to today’s Council on Foreign Relations meeting with Michael S. Barr. He is the vice chair for supervision, currently, of the Federal Reserve Bank Board of Governors. Recently announced that he’s going to be stepping down from that role, which we’re going to talk about, but that he is going to remain a governor at the Federal Reserve.

This meeting is part of the C. Peter McColough Series on International Economics. I’m Andrew Ross Sorkin. I’m a columnist at the New York Times. I’m a coanchor of Squawk Box on CNBC. And I’m going to be presiding over today’s discussion. We are joined today by CFR members attending in person and also, of course, this event is being streamed online.

Please join me in welcoming Vice Chair Barr to the podium. He’s going to give some opening remarks before we get into our discussion. (Applause.)

BARR: We’re still working on the choreography. It’s great to see everybody. Thanks for being here today. I’m really looking forward to the conversation with Andrew. I’m going to make some formal remarks first. And in my remarks, as you see from the announcement, I’m going to be focusing on artificial intelligence.

Advances in artificial intelligence have accelerated rapidly over the past few years. It is now commonplace to see autonomous vehicles navigating city streets and generative AI tools are available on phones and other devices wherever we go. AI innovation makes headlines and plays a big role in financial markets and generative AI has the potential to change how we think about productivity, labor markets, and the macro economy. Today I will address that question by outlining two hypothetical scenarios for AI’s impact and the implications for businesses, regulators, and society. I will focus my comments on generative AI, or gen AI, a subject of AI that has seen significant growth and integration into economic activity in just a few short years.

Compared to earlier iterations of AI, gen AI is able to generate content which allows it to significantly enhance productivity across a range of knowledge-based activities and to be used by people without coding skills. Gen AI will likely become, in my judgment, a general-purpose technology with widespread adoption, continuous improvement, and productivity enhancements to a wide range of sectors across the economy. We’re already seeing gen AI improve the productivity of its own R&D. There is widespread enthusiasm for gen AI and survey evidence shows much faster rates of consumer adoption of gen AI than were seen for personal computers or the internet.

While actual deployment of gen AI is limited to some business functions, and there have been pitfalls along the way, businesses in almost every sector are experimenting with or considering how to make use of the technology. Firms are also exploring agentic AI, gen AI systems that not only produce new content but are also able to proactively pursue goals by generative innovating solutions and acting upon them at speed and scale. Imagining agentic AI’s ultimate application, some speculate that we could experience, quote, “a country of geniuses in a datacenter,” unquote, a collective intelligence that surpasses human capabilities in problem solving and collaboration. Some believe agentic AI has the potential to connect ideas in disparate domains, potentially transforming research and development and society more broadly.

Today, I will outline two hypothetical scenarios for considering how gen AI could evolve. In one, we see only incremental adaptation that primarily augments what humans do today but still leads to widespread productivity gains. In the other, we see transformative change where we extend human capabilities with far-reaching consequences. For each scenario I consider the potential implications for the economy and the financial sector. Thinking through hypothetical scenarios can help us widen our lens to a range of possible outcomes and provide a framework for assessing the balance between benefits and risks. Scenarios are not predictions of the future, but provide a framework for analyzing the factors that could lead to different outcomes. Reality is complex. Gen AI adoption rates will vary across industries, leading to diverse impacts on market structures. Elements of both scenarios will likely come to pass and play out at different rates, which will influence the effects on the economy and society.

In the short term, gen AI may be overhyped, while in the long run it may be underappreciated. And, of course, things might turn out differently from these hypotheticals. So let me focus on hypothetical one. Let me begin with the incremental scenario, where gen AI primarily augments work and existing processes and leads to steady and widespread productivity gains but does not fundamentally unlock new capabilities or transform the economy. In this state of the world, gen AI tools enhance efficiency and enable more personalized solutions across industries in ways that have incremental but still meaningful effect on people’s lives. For instance, in customer service, professional writing—but not this speech—(laughter)—and software engineering, gen AI-powered tools are already supporting workers, improving accuracy and speed. And these effects could spread to other sectors.

In this world, healthcare sees significant improvements as gen AI reduces administrative burdens, assists with diagnostics, and personalizes treatment plans based on real-time patient data. Medicines and other treatments are developed at a faster pace. Education is similarly affected as gen AI alleviates administrative tasks for teachers, allows lessons to be tailored to individual students, and permits students to learn by doing. In manufacturing, gen AI optimize supply chains, anticipate and adjust more quickly to disruptions. And current manufacturing processes are refined through virtual iteration. In materials science, gen AI-driven experimentation accelerates the discovery of new material, leading to advances in everything from construction to electronics.

Turning to the financial sector, we could see similar productivity gains. Community banks leverage gen AI-powered chatbots to provide customized financial advice rooted in local knowledge, while institutions of all sizes continue to advance use of gen AI for compliance monitoring, fraud detection, risk management and document analysis. The impact to society would be incrementally positive in this state of the world. Humans could use gen AI as a tool to deliver goods and services that we currently produce in a more efficient way. Productivity would go up. The economy would grow at a faster pace.

What does this mean for the labor force? The impact will depend on the industry and the nature of the job. Gen AI experiments suggest that technology holds the promise of leveling up skills and bringing productivity of lower-performing workers into line with higher-performing workers. In other cases, it could augment the highest performers, leaving them more time for creativity or strategic aspect of their roles. Increasing automation for certain tasks may displace some workers, where certain skills can be replicated by gen AI. Historically, as technology has replaced some jobs it has augmented existing roles or created new ones. However, this is not to downplay the individual cost for workers who need to retrain, find other employment, or change careers in response to major changes in labor demand. Society will need to account for these possible effects of AI.

What does it mean for the economy? As I noted before, the economy should grow if the incremental productivity gains are widespread. However, in this scenario, it is possible that the expected value creation from gen AI was overhyped, anticipating transformative breakthroughs rather than incremental productivity gains. This could trigger market corrections for the firms that have heavily invested in this technology if reality doesn’t measure up to expectations. While the U.S. economy experienced a surge of productivity growth during the dot-com boom in the late 1990s, it was followed by a wave of bankruptcies, capital overhang, and a cautious business investment climate. The effects of the ensuing recession were widespread.

What does this mean for financial stability and other financial risks? In this incremental scenario, gen AI may magnify both the vulnerabilities and sources of resilience that already exist in the system. Attractive trades become more crowded, but risk managers gain new insights. Malicious actors gain new tools, but cyber defenders become better armed. So long as financial regulators, enterprise risk managers, and others charged with managing these downside risks prioritize efforts to keep pace with the evolving financial ecosystem, there’s nothing to suggest a wholesale transformation of the balance of risks. Of course, keeping pace will pose challenges. And it’s important that we all focus on the need to meet these risks.

Now, let’s consider a more dramatic hypothetical scenario, scenario number two, in which gen AI adoption extends beyond improving what we already do and provides new expertise and capabilities that have transformative effects on the economy and society. In this scenario, humans deploy imagination and creativity, combined with robust investment in research and development, to deploy intelligent gen AI systems to make rapid breakthroughs in, for example, biotechnology, robotics, and energy, fundamentally reshaping existing industries and creating new ones. In this instance, to focus the mind, we can think of gen AI as no longer a tool for scientists to analyze data. In a sense, as some have said, it becomes the scientists directing the research.

For instance, let’s say gen AI applications in healthcare do not simply improve how we currently deliver care, but also, for instance, enable therapies that target genetic mutations and cure diseases previously considered incurable. Similarly, manufacturing evolves to create gen AI-driven robotic factories with goods produced with new materials and atomic precision. Materials science is transformed through the discovery of programmable materials and self-healing substances, all of which reshape construction, technology, and consumer goods. Meanwhile, gen AI optimizes fusion energy research, expediting the shift to sustainable energy sources. And gen AI helps to create the next generation of quantum computing. In that way, gen AI improves its own energy sources and computing capabilities, enabling it to become a more powerful creative tool.

Finance also looks radically different than it does today in this scenario. Individuals with access to hyper-personalized financial planning and businesses with innovative products and services seamlessly connect with one another through near frictionless or novel forms of financial intermediation. Trading strategies and risk management practices are boosted by greater gen AI-based analytic tools that have dynamic, real-time access to an enormous knowledge base in both public and private domains. Although this transformative scenario may be more speculative, and is accomplished—it accompanied by a far greater degree of uncertainty than the first, it is important to consider, given the extraordinary opportunities for human advancement and welfare that could arise even if just one of its transformative components were to come to fruition.

We would need to fundamentally reimagine how the economy is structured. What are the impacts on the labor force in a world where gen AI’s capabilities extend beyond what humans can accomplish today? Humans may have a role to manage multi-agent gen AI frameworks or fill gaps where gen AI solutions remain expensive or inefficient for some applications. But this is a world where some workers may see their current jobs disappearing, and in which they may also see their own work transformed and may have many more choices about the work that they do. The nature of labor would radically change. And this would require us to have broader conversations about how to organize the economy. These conversations should wrestle with how to navigate major economic shifts in a way that recognizes the impact on the human condition and the extent to which people derive their communities, friendships, personal sense of meaning, and dignity from their work.

What about the competitive landscape? There’s probably a greater likelihood that rewards for businesses would be distributed more unevenly at first, as significant breakthroughs with far-reaching ramifications may benefit a subset of firms and industries, and concentrate economic power in firms that control gen AI breakthroughs. If only a handful of firms have the ability to accomplish the incredible things I’ve mentioned above, they may dominate markets and crowd out competitors. However, to the extent that gen AI becomes broadly effective, widely available, and cheap, these market advantages could lessen over time if the right regulatory environment supports competitive market dynamics. But history suggests caution in this regard. A handful of players may dominate.

And finally, for finance, we should anticipate fundamental changes in this scenario. When it’s working well the financial system helps move money and risk through time and space. To the extent that there are fundamental changes to how the economy is organized, we could see a new set of institutions, markets, and products to facilitate transactions among households, businesses, and gen AI agents. So what should we do? Among the many ways in which we can help to harness the potential benefits of gen AI and minimize its risks, I highlight only a couple today. Financial institutions and the Federal Reserve system should consider investing sufficient resources in understanding gen AI technology, incorporating it into their workflows where appropriate, and training staff on how to use the technology responsibly and effectively. Meanwhile, the financial regulatory community should approach the changing landscape with agility and flexibility.

And beyond the financial sector, collaboration between government, private industry, and research institutions will be critical to ensure that gen AI systems are not weaponized in catastrophic ways. We should continue to focus on responsible AI research and development and implement safeguards against misuse, including monitoring systems, standards for secure AI systems development, and agreement on red lines for acceptable use cases. We should be attuned to the impact of gen AI on our economic and political institutions. There’s a risk that it concentrates economic and political power in the hands of the very few and could lead to the gains being realized by only a small group, while the rest are left behind.

Another area I want to emphasize is AI governance. I think most would agree that the goal of the technology is to improve the human condition. And to do that, we need to be intentional in advancing that goal. We should make sure that we think about gen AI as enhancing not replacing humans, and set up best practices and cultural norms to that end. Every financial institution should recognize the limitations of the technology, explore where and when gen AI belongs in any process, and identify how humans can be best positioned in the loop. We should also focus on data quality and make sure that gen AI uses do not perpetuate or amplify biases inherent in the data used to train the system or make incorrect inferences to the extent that the data is incomplete or nonrepresentative.

In the realm of regulation, frameworks for understanding model risk may need to be updated to address the complexity and challenges of explaining AI methods and the difficulty of assessing data quality. We need to be attuned to the risks in finance. The very attributes that make gen AI attractive—the speed, automaticity, and ability to optimize financial strategies—also present risks. When the technology becomes ubiquitous, use of gen AI could lead to herding behavior and the concentration of risk, potentially amplifying market volatility. As they will be directed to maximize profit, gen AI agents may converge on strategies to maximize returns through coordinated market manipulation, potentially fueling asset bubbles and crashes. Speed, automaticity, and ubiquity could generate new risks at wide scale.

We also should monitor how introduction of this technology changes the banking landscape. Nonbanks may be more nimble and risk forward in incorporated gen AI into their operations, which may push intermediation to less regulated, less transparent corners of the financial sector. In addition, this competitive pressure may push all institutions, including regulated institutions, to take a more aggressive approach to gen AI adoption, heightening the governance alignment and financial risks I mentioned above.

In conclusion, while AI’s impact will vary across industries, and the reality is evolving, the scenarios I’ve outlined today provide a framework to begin thinking about how we should respond to developments in gen AI. However, as I mentioned above, elements of both scenarios will likely be present in the future and play out at different rates, which will influence the effects on the economy and society. Rapid advances in this technology, such as agentic AI, and advancements in open-source models underscore just how new this technology is and the importance of understanding what it means for individuals, businesses, and markets. Thank you very much. (Applause.)

SORKIN: Thank you for that. I have about a million questions for you. (Laughter.) And we are all drinking out of a firehose of news, so we’re going to get to a whole bunch of things. But I do have one major follow up just on the conversation that you proposed around AI. Which is, we’ve heard from people like Sam Altman, who runs OpenAI, that the employment picture—and you touched on employment, but only this much. He talks about the need, potentially, for UBI, universal basic income. That employment could be so devastated, maybe by 10, 20, 30, 40 percent literally pulled out of the market, even if it’s pulled out of the market for some temporary period of time. How do you contemplate the true employment implications of all of this?

BARR: You know, Andrew, it’s a great question. I mean, part of the reason that I used the scenario-based approach—and it was just two scenarios among many possible narratives—

SORKIN: And which one are you betting on, by the way?

BARR: Is that we don’t know. (Laughter.) Like, we don’t know exactly, you know, how this is all going to turn out. We have seen lots of episodes throughout history of technological innovation and people being concerned that the new technology was going to make everything that currently existed obsolete. And what we’ve seen most of the time in history is that the labor market adjusts, but it is a very painful adjustment. So the kinds of things that we need to be focused on are things like universal basic income, income supports, earned income tax credit, retraining, measures, all those things. But it is likely that this technology is going to be very disruptive, and we don’t know precisely in what ways. I think it’s the right kind of question that we should be addressing now because, as I said, I mean, my view is that AI is kind of overhyped right now, in the short term. I think people are expecting more from it last year, this year, next year. But, I think, probably underappreciated the long term, in terms of the kinds of changes that it could cause in our society. So the right time to ask is now.

SORKIN: I said there were a million headlines, and I want to talk about them. One of those headlines was you. So I want to go there, if we could, which is that on January 6 you said that you were going to resign as the vice chair of supervision from the Fed. Take us inside the room in terms of your decision to do this. Obviously, there was a lot of speculation, and I have to imagine—I know it was part of your thought process, that President Trump was going to push you, and you jumped before he pushed you. But walk us through the whole process.

BARR: Yeah, Andrew, you know, it’s—it was a very, very difficult decision for me to make. And I’d say, in some ways, a very painful decision. But what I did is I just looked at the cold, hard facts. And my judgment was that if I stayed in the role that the fight over this role would become a major distraction for the Federal Reserve. It would be a distraction for the Fed, it’d be a distraction for me, it’d be a distraction for the team of people that I lead. And that distraction would have prevented us from doing our job serving the country. And to me, it wasn’t—just wasn’t worth it. So I decided to step down from the vice chair role, continue on as governor, where I’ll remain. My term goes to 2032, which is a long time from now. And, you know, I figured I could best serve the public by staying on in that role and not becoming a distraction from our work.

SORKIN: OK, but all of this is in the context of the independence of the Federal Reserve. And I want to read you something. This is Natasha Sarin. This was an op-ed in the Washington Post. I’m sure you read it. This is in the context of what—of your decision to resign. “If the Federal Reserve is politicized and weakened, the winners will be large financial institutions. The losers will be all of us. Unfortunately, that’s already happening before President-elect Donald Trump is even sworn in. Choosing not to fight the battle is the same thing as losing it. The Federal Reserve is signaling to banks that the sheriff has left town, somewhat literally in Barr’s case.”

BARR: Hey, look, I said this is—this was a difficult decision for me to make, and personally painful to make. But I’m the only one who can make it. And unless you’ve sat in the chair, and you’ve looked at the tradeoffs, it’s very easy to be a commentator on the sidelines and take shots. I’m used to people taking shots. It doesn’t bother me. I think I made the right thing—the right judgment for the Fed, the right judgment for our economy, the right judgment overall for serving the public. Other people can have different views. I understand that. If it were an easy call, it would be an easy call to make. (Laughs.) It’s not an easy call.

SORKIN: But because—

BARR: But I think that that protected us.

SORKIN: I mean, you made the decision because you felt that the independence of the Fed would otherwise come under even more intense pressure.

BARR: Yeah. I mean, I—you know, my judgment was that fighting the fight about this position would be a huge distraction to the Fed. It would put the Fed in the crosshairs. And that’s not a place that I think was good for the Fed to be. And so I made the judgment to step down. Again, I recognize that other people could make other judgments. But I think that’s the right judgment to have made. And I feel comfortable with it.

SORKIN: Let me ask you a separate question. It’s somewhat related, which is, given your role supervising banks, one of the other institutions that is being dismantled, quite literally, in front of us is the CFPB. What do you make of that? Are you concerned about that?

BARR: You know, let me just, you know, take a step back. It is very easy in the current environment we’re in to be focused on the kind of day-to-day issues that arise. And it’s totally appropriate to focus on those things. They’re meaningful. They’re real things. But if you take a step back and look at the longer sweep of history, we have very, very strong institutions in the United States. They have survived lots of different stresses to them over the sweep of our history. We have an enormously successful, vibrant, energized economy that survives the ebbs and flows of different administrations. So we’re an incredibly strong country, big-picture sense. And I’m not worried in that sense, about the country.

SORKIN: OK. And we’re going to talk about the economy and where you think we are. But let me—let me ask you then another headline, which maybe you’ll disregard but, nonetheless, I think a lot of people are concerned or at least questioning it. I see Jack Lew is in our audience. He signed a letter, along with a whole number of other former Treasury secretaries very concerned about the payment system in America. Our current Treasury Secretary Scott Bessent has said that the DOGE team is not, in fact, inside of the Federal Reserve, but is in fact inside the Treasury. But that payments can’t really be made from the Treasury. Speak to this. What do you make of what is happening here?

BARR: Well, you know, let me say, first of all, I appreciated very much Secretary Lew’s letter, and the letter signed by other former treasury secretaries, on this issue. It’s really important that we have a payment system that the American people can rely on. When Congress passes laws, agencies make payments according to those laws. The payments should flow through to the people they’re supposed to. You know, if I’m due a tax refund, I expect that my tax refund will be paid to me according to the law. So we need to make sure we have a system with integrity and, you know, that’s absolutely—

SORKIN: So then the question is, does it have that integrity? Given that Scott Bessent has said that the payments are made by the Federal Reserve and actually not by Treasury, is he accurate about that?

BARR: We receive—the Federal Reserve banks act as agents for the Treasury Department in making payments for the United States government. So any file that we receive from the Treasury Department we process and we pay according to the law. And so that’s always been the case. And the Federal Reserve is doing that now. I can’t speak to anything that the United States government decides to do, but once the government—once the Treasury sends us a file we make the payments on that according to the law.

SORKIN: Let me ask you maybe slightly differently. If you were put in charge of trying to find efficiencies in the government, and it hasn’t been done this way before, but would you go to the Treasury and try to get inside the systems, because, you know, like a great bank robber, they always say you got to go with the money is. You got to go to the bank. Is this where you go to try to understand where all the money is moving? Is this the right way to approach it, or do you think it’s not?

BARR: Well, no, let me—I can’t really speak to the motivations of the folks doing it. Let me just say that the Federal Reserve’s role in this is a very defined role. We receive files from the Treasury Department. We pay through the payment system on those files. It’s the way we’ve always done it. People can have confidence that we’re doing it that way too.

SORKIN: But we don’t know if the payments are being changed on the other end.

BARR: I don’t have anything at all to say about that. (Laughs.) I don’t know anything about it.

SORKIN: Let me ask you a little bit about where we are in this economy right now. We just had CPI numbers that came out last week, hotter than expected, about 4 ½ percent. That is still far from the 2 percent goal that the Fed would like to see. Where do you think we really are?

BARR: So, you know, I think we’ve made a lot of progress on inflation. We’ve got, first of all, a strong economy. We had really strong growth last year, over 2 percent—nearly 2 ½ percent growth last year. We’ve got an unemployment rate that is low, at about 4 percent. It seems stable and has been stable since last summer. That’s very encouraging. And if you look at inflation, you know, we think of it in PCE terms. It’s gone from over 7 percent headline, you know, down to 2.6 percent. It’s gone from 5.8 percent to 2.8 percent in—thereabouts, in core terms. So we’ve made enormous progress on inflation. There’s still a lot more we need to do, but the overall path of inflation has been on, I would say, a bumpy, but steady and I think sustained, path towards 2 percent.

The bump up that we saw in January was something that we anticipated based on the seasonal pattern that we’ve seen in the past. I expect it will moderate again in February and March. So that’s all, I think, positive. And then, you know, we need to look at what the policy environment is and how the policy environment may change the trajectory of all of that.

SORKIN: Well, so let me ask you about that. Which is the other sort of main headline that we talk about almost every day endlessly is tariffs, and how you think the Fed should be thinking about that. I will tell you that Christopher Waller just said that his baseline view—this is actually new just today—is that any imposition of tariffs will only modestly increase prices, and in a nonpersistent manner. Is that a widely held view?

BARR: Well, I don’t want to speak for Governor Waller or for any of my colleagues. I’ll just tell you what I think. I think it’s too early to know the answer to that. So there’s a lot of uncertainty about the policy path for the government. There’s uncertainty about tariffs, who they’re going to apply to, how long they’re going to apply, for what kind of products, how they affect final goods and intermediate goods, you know, the extent to which they may be permanent or may be temporary. All that’s uncertain. And there’s uncertainty on the fiscal policy path more broadly in terms of tax cuts. There’s uncertainty about this shape and scope of immigration policy. So at least for me, I’m in a, you know, wait and see position. I’d like to understand better what the policy is, and then decide what the implications for the economy might be, and then decide what the implications for monetary policy might be. So I’m not—I’m not in a rush to answer that question.

SORKIN: OK. What do you make of the argument that, by default, that means you’re going to be late?

BARR: I don’t think so. I mean, look, we’re—you know, all of us, in our individual way, make our own judgments about this. But we go meeting by meeting and we’re looking at the data as it evolves. And right now, if you look out at the economy, the economy is strong. It’s been a strong economy for quite some time. The unemployment rate was ticking up last year. And that caused me some concern. But it then moderated and has stayed pretty steady. And so the unemployment rate feels stable. The labor situation feels stable. The labor market has cooled a great deal—(audio break)—actually, I think, take the time we need. Now, if it turns out that inflation drops faster than we expected, or if it turns out the labor market weakens more than expected, then we can act. I think we’re in a position to do that.

SORKIN: As it relates to immigration and this crackdown, how quickly do you think we’re going to see that? And how quickly do you think we’re going to see this in the numbers? We’re not at all.

BARR: So, you know, the immigration policy itself is up to the administration. I don’t have, you know, a comment on what they decide to do. And we take that in our forecasts. And to me, it’s just too early to know. It’s too early to know what the policy will be. It’s too early to know what the effects of the policy will be on the economy, and therefore what the implications, if any, for monetary policy are. You know, as a general matter, if you reduce the labor force, you reduce potential output, and you reduce output. So we’ll look to see whether that is of a scale to see that in the economy.

SORKIN: Right. Curious, maybe this goes back to a supervision question, but there’s been a lot of headlines recently about this idea of debanking. In fact, Jay Powell was asked about it just last week. And whether, in fact, you believe that banks in the United States have effectively debanked certain customers and clients because of their political views. In fact, President Trump, very directly said to Brian Moynihan, who runs the Bank of America, that he believed that Bank of America, and that Jamie Dimon at JPMorgan, was engaged in debanking.

SORKIN: You know, I haven’t seen that kind of—I haven’t seen any evidence of that kind of activity. You know, the thing that people raise more directly is, is a bank engaged in activities that are higher risk and therefore reducing their client exposure because of concerns about money laundering, or terrorist financing, for example. And that, of course, happens. If you’re a bank and you’re worried that you have a client that may be engaged in unlawful activity, you’re going to reduce your exposure to that kind of risk. That’s sort of appropriate risk management. But I haven’t seen this—just my own view—I haven’t seen the kind of thing that you described.

SORKIN: Right. Jamie Dimon, apparently, made reference to this in Washington, and suggested that actually the regulations that are put on the banks in some ways have required the banks to not do business with certain folks. And that because of the kinds of industries that you may be referring to, but also I’d include crypto in that, that in fact they have had to debank certain clients. Not because they say they want to, but because they say that some of—but this may relate to some of the supervisory issues that have come out of Washington. And obviously you’ve been in that role.

BARR: Andrew, you know, we try very much at the Fed to kind of play straight up on the middle on these issues. So we have institutions, banks of all sizes, supervised by the Fed that are engaged in crypto-related activities—either themselves directly or supporting clients who do crypto work. Our expectation is they do that in a safe and sound way, and in compliance with the consumer laws, and in compliance with laws against money laundering and terrorist financing. And if they do that, that’s fine with us. We have a novel supervision program designed to supervise institutions in this space to make sure we have the expertise we need to do that right. And we have institutions doing it.

And then we have other institutions that did it wrong, and got themselves into trouble either because of AML/BSA problems or because they were unable to meet their liquidity needs of their clients, and they went out of business. So to me, at least, it’s just straight up the middle risk management and banking. If you do it right you can do it, and if you don’t do it right then you shouldn’t be doing it.

SORKIN: I do want to open up to questions in one moment, but just to bring this all back to AI. I’m very curious if you use AI inside the Federal Reserve.

BARR: We do. We’re very, very cautious, as you’d expect, at the Fed—(laughs)—in using AI. We have a very strong system of internal governance over it. But we experiment with using it in very controlled ways. So, you know, for example, we have a lot of coders at the Fed, because we run the payment system. We have thousands of engineers at the Federal Reserve. And one of the things, as you know, coders have to do is test their code. And the unit testing you do for code testing is pretty boring work, and it takes a long time. So we’ve experimented with using generative AI, along with a human coder, to see if we could speed up that process. And, yes, we could. (Laughs.) So those are the kinds of efficiency improvements that I think you’ll see, you know, more broadly in society. So, yes, we use it, but we are super careful about it, and with long-term controls.

SORKIN: Do you think—do you think—you have—you have several hundred economists who work with the Fed.

BARR: Mmm hmm.

SORKIN: Do you think long term you will need all of those economists?

BARR: I think we might need more of them. (Laughs.) You know, the economy is super complicated. We use them—you know, we use models today. A lot of what the Federal Reserve economist do is understand models. And we take those models as an input into a forecast. But the forecast involves a lot of human judgment. You know, we were just talking before about the uncertainty about policy. We have to make a decision how much of that uncertainty to build into the forecast or not. Human judgment is an important part of that. And models are really good at data-rich environments, and less good at places where there’s less historical experience. And so you have to also account for that. I think economists, and engineers, and coders, and writers, and philosophers, everybody else, will still have plenty of work.

SORKIN: I hope you’re right. Let’s open it up.

BARR: And certainly journalists.

SORKIN: Let’s open up for questions. We’ll go to the front and we’ll make our way back.

Q: Thank you very much. Rebecca Patterson. I’m a senior fellow here at CFR.

There’s so many questions I’d like to ask, but I’d like to stay in the supervisory part of your world. President Trump is encouraging development of stablecoins to have a digital dollar that could circulate globally, provide a new source, or a greater source of demand for U.S. Treasurys, at a time when our supply is increasing. What type of—do you think we will have the regulatory framework we need around stablecoins that will make them truly stable? Tether, for example, right now, about 15 percent of its reserves are not Treasury or cash, like securities. And do you think it’s necessary for us to prevent further research on a central bank digital currency at the same time? That’s what President Trump has said in his executive order, that that research is basically done.

BARR: So on the first, we don’t currently have any stablecoin regulatory arrangement at all. And it would be good for us to have a stablecoin arrangement in the United States. The Federal Reserve has been working, at the request of Congress, for many years now to try and develop a consensus framework in the United States that we could support. We’ll continue to work on that. There was a lot of effort made on a bipartisan basis over the last three years, or more, to develop that kind of consensus approach for stablecoin legislation. I was very encouraged by that process of bipartisanship, and I hope that—I hope that it continues, and that we can continue to offer our expertise for that process.

You know, stablecoins—dollar-based stablecoins really, you know, borrow the trust of the Federal Reserve. And so it’s absolutely essential for us that there be meaningful regulation and supervision of stablecoin issuers. And I would very strongly support that. Whether stablecoins play this broader role or not in the economy I think is a separate question. If we have material amounts of stablecoin issuance in the country, we want to make sure that they’re done in a safe and sound way.

Central Bank digital currency conversation is a different conversation. We’ve said at the Fed very clearly that whether we issue a retail CBDC is—that’s a matter for the president and for the Congress. That’s not something the Fed could just go do, or would we ever want to go do. That’s for the elected branches to decide what to do.

SORKIN: Question up here, and then we’ll get there.

Q: Great. Thank you, Andrew. Joseph Gasparro, Royal Bank of Canada. Thank you, again, for all of your service.

How do you think about broader solidarity or regulatory harmony when it comes to AI governance? If you think of a case study for GDPR, where the U.S. had to import this European regulation and potentially it kind of, you know, curtailed or slowed down various industries in the U.S.? Thank you.

BARR: You know, it’s very early in the process to think about AI regulation. I think it’s the right time to have that conversation. And I do think the discussions we have across borders make a lot of sense. They add to our ability to make good judgments about it. But I’m not sure that it’s a space where we need to reach 100 percent agreement on an approach across jurisdictions. It’s, I think, new enough that a little bit of give in the joints and a little bit of experimentation actually might be—might be a positive development. But I feel that way only in the short term. That is, in the short term I think we have the space to figure this out, and get this right, and be flexible in the joints, so that over the longer term we have a framework get in place that’s protective of the right set of values, and we’re not talking in ten or twenty years about the kind of downside scenarios I discussed before.

SORKIN: Ma’am. And then I’ll come over.

Q: Hi. I’m a term member here at CFR. And I was illegally fired from the CFPB last week.

I led the CFPB’s work monitoring consumer financial markets’ use of AI. And my question is, what does the financial marketplace look like without a functioning CFPB?

BARR: Look, I think it’s really important that we have significant oversight of the financial sector for both safety and soundness, and for consumer protection. The CFPB, as you know, and many members of the audience know, has jurisdiction both for the largest institutions in the country for consumer compliance and for the whole nonbank financial sector, which is not—both of—neither of those areas are covered by the Federal Reserve or other prudential regulators. So I think, you know, a CFPB on the block is an important part of the overall picture we need to make sure we have a marketplace that is appropriately covered for all products and services that affect consumers.

SORKIN: Question back here, and then we’ll move over there.

Q: Vice Chair Barr, thank you for your comments. Andrew Watrous with Morgan Stanley.

I have a question about the GSEs. And I’m curious how you see the regulatory environment for the GSEs evolving in a world where they’re privatized. And I’m especially interested in the risk weights that could be applied to GSE securities. Thank you.

BARR: So, you know, the oversight of the GSEs has been a hot political topic for many administrations at this point. (Laughs.) And I won’t prognosticate on whether that will be the same or change under the—under the new administration. It’s sort of far outside my lane. We kind of take the structure of the mortgage market as a given as an input into how we think about the economy and how we think about capital rules, and the like. And I just don’t want to speculate about the direction—(off mic).

SORKIN: Ma’am.

Q: Hello. Thank you. Alma Caballero. I’m a term member here.

Vice Chair, you mentioned that you’re not worried about the Fed’s independence and that you’re confident in the strength of the U.S. institutions. Where do you get that certainty from?

BARR: I’m not sure I said any of the exact words that you just said. So let me just say what I think. (Laughs.) I think overall that the strength of U.S. institutions—that the U.S. institutions are strong. And then if you look over the long sweep of history, they’ve stood the test of time. I think that’s true of the Federal Reserve too. You know, we’ve been an agency, an independent agency, for more than a hundred years. And that has served the American people well. And so I’m really talking about in the large historical sense. It’s not—I was not giving a don’t worry, be happy speech.

SORKIN: Can you—by the way, can you recall a time in history where you think that the Fed was—the Fed’s independence was under more question?

BARR: Well, you know, I think if you look, again, at the long sweep of history, the creation of the First Bank of the United States was a deeply politically fraught moment in the United States history. And it was contended with for its whole sweep and allowed to end. And then we had the Second Bank of the United States. That was also a very contentious political moment the United States, and it was allowed to end. And then we had the creation of the Federal Reserve, which was, as you know, a compromise among competing political forces in our country, between people who wanted a powerful Central Bank in Washington and people who wanted it distributed around the country, between people who wanted a public institution and people wanted a private institution. And we’re that compromise. So it’s not unusual for the country to be fighting about the role of the central bank. Again, I think if you have a long enough sweep of history it helps, you know, set the conversation a little bit in perspective.

SORKIN: (Inaudible)—had a question, then we’ll go to you. And then we’ll—I think there’s a question there too.

Q: One thing I’d like to ask you about, that you touched on briefly in terms of what you’re looking at in terms of the future of—the structure of the economy, and that’s the debt situation. And without going into all the details, we’re going to have clearly a very heated debate on the debt issue coming up very shortly. I wonder if you could give us your thoughts on what the dangers are of the trajectory that we now see and is sort of validated by a lot of economists. And it’s hard to know when the problem will occur, but we know that when these things happen, at some point a problem does occur. I wonder if you could elaborate on your thoughts on when we approach the danger zone, and what the dangers you see are of the sort of broken fiscal system that we seem to have had over the last couple of decades.

BARR: Yes. So let me—let me do this in as careful a way as I can. So I don’t want to comment at all on the current fiscal conversation. That is, what the appropriate mix of spending and revenue is, and what choices Congress should make. That’s for the president and the Congress. But taking a longer-term perspective, the trajectory of debt to GDP keeps going up and up. It varies a little bit on whose model you use, but the direction is not in in doubt. And most people would say that that’s not a sustainable path for the debt of the U.S., or any economy. And what they mean by that is a bunch of different things. One is that the portion of the expenditures that we all engage in on an annual basis become increasingly apportioned to paying off that debt, the interest expense becomes a larger and larger portion of the budget, which crowds out other things we might want to do as a society.

It shifts the burden from our generation to our children’s generation and grandchildren’s generation. It probably increases the term premium that you pay, that the federal government pays, to issue debt, which means it’s even more expensive over time to issue debt. And other countries have seen, at very, very high levels, that, you know, not—you know, not—there’s no uniform rule, but at very high levels it becomes—the financial system becomes more fragile and more susceptible to shock. So it is, in a long-range sense, something to worry a lot about. And of course, the longer you watch it, the harder it becomes to adjust.

SORKIN: Sir. We’ll get you a mic.

Q: My name is Andrew Gundlach, CEO of Bleichroeder.

I wanted to follow up on Andrew’s question about tariffs. As you know, they start with pretty weak currencies relative to the dollar around the world. And if you can share with us your thoughts about the pluses and minuses of resetting the terms of trade bilaterally through tariffs, as is currently being done, as opposed to multilaterally, through a revaluation of certain overseas currencies, Plaza Accord type of thing, and whether or not you think, ultimately, if the goal is either to change the terms of trade or revalue the dollar—Trump has not said which one it is, nor has Scott yet—if we’re necessarily going to get to a Plaza-type two Accord. Which, to Andrew’s question, would certainly be, at least theoretically, inflationary, based on histories.

BARR: So it’s a really terrific question, but also one I’m going to be really careful not to answer. (Laughter.) So like my answer on fiscal policy, let me not get into the short-term tariff debate. At the Fed, we take the government’s decisions about that as given. That is, the president and the executive branch and Congress get to decide what our tariff policy is. And we’ll take those into account in our forecasts of what’s happening in the economy and then decide how to react in terms of monetary policy. You know, tariffs, just as a general matter, you know, will raise the price level. And the question is, do they raise the price level on a short-term basis? Do inflation expectations remain well anchored, or not? How much persistence is there in price increases because of the effects on intermediate goods? It might play out over time. Those are the kinds of questions that we ask ourselves in thinking about the forecast. But we don’t weigh into the question of whether we should have tariffs or shouldn’t, or what they should be framing.

SORKIN: Can I actually just follow with this? Because this is something actually we do know. So one of the things that is happening right now, even though we don’t know what the tariffs will be permanently or even on a short-term basis, we hear at least anecdotally from folks like the CEO of Ford, who says that right now that company is going back and forth between the border of the U.S. and Mexico, bringing products and supplies to the United States as quickly as humanly possible, at a cost to them today, because they don’t know the uncertainty on the other side of what’s going to happen. That has not emerged in the numbers yet, but when you think about that, what does that what—how do you assess that situation? And, given all the conversations I’m sure you’re having with business leaders and others today, putting aside what the tariffs are in the future, just the prospect of the tariffs are doing, what?

BARR: So the short answer is—it’s consistent with what you just said—it’s not showing up the macro data yet. So we’re not seeing the kinds of shifts that would suggest widespread activity of the kind that you described, although in discussions with business leaders it’s clear that it is going on in some instances. And if you think about automobile supply chain, you know, the parts that are used in an automobile don’t just come into a country or out of a country. In our case, they go in and out of Canada, Mexico, the United States, as the car is being built, multiple times. So manufacturers are thinking about how to organize their supply chains given that.

It’s very hard to change your supply chain in a short period of time. So there will be price changes within those products. And then we’ll be looking at how those flow through both, you know, at the border in terms of import prices, but also into intermediate goods, and then into final consumer demand. And it’s too early to know the answer to the question of whether it will be macro economically significant or not. I don’t know. But directionally, you know, tariffs have the effect of increasing prices, reducing productivity, reducing growth. And you need to look at how those things offset each other to understand whether or not there should be a monetary policy reaction function to that basket of changes.

SORKIN: Right. We are almost out of time and I know there’s been a question in the corner in the back for a long time. Thank you for your patience.

Q: I’m Ken Kraetzer with CaMMVets Media.

I just want to go back to Mr. Barr’s opening about AI. I worked in banks for a long time in my career. Does AI increase the risk of cyber threats to banks? And if you’re advising banks, would you say that AI is a tool to help prevent cyberattacks or to detect them, like the military does? Thank you.

BARR: It’s a great question. I think, in a sense, the answer to your question is both. That is, AI is increasing the ability of fraudsters and cyberattackers to infiltrate the financial system. And it is increasing the ability of banks to protect against infiltration by those attackers. And there’s going to be kind of an arms race about use of AI for good and for ill in that space. And so I do think that banks are, and should be, investing in additional skills for their workers, additional R&D for the technology, as they are, to make sure that they stay current, or they can get ahead of that arms race. But it’s happening and it will happen.

SORKIN: And before we let you go, do you know who is going to take your supervisory role? Who President Trump is going to appoint?

BARR: No. I do not know.

SORKIN: Do you know if he’s going to appoint anybody?

BARR: I do not know.

SORKIN: Have you had any conversations with anybody about this?

BARR: No. (Laughter.) It’s anybody’s guess.

SORKIN: Vice Chair Barr, thank you very, very much.

BARR: Thank you, Andrew.

SORKIN: Appreciate it. Thank you. Thank you for the questions. (Applause.)

BARR: It’s been a pleasure.

(END)

This is an uncorrected transcript.