President, Federal Reserve Bank of San Francisco
Fed President Mary Daly on the AI Economic Reality Check
#businessnews #stockmarketnews #financialnews
Mary Daly, President of the San Francisco Fed, offers her perspective on the economic news surrounding artificial intelligence. She notes that while there's significant investment in ai, tangible productivity gains are not yet evident in economic data. Daly also addresses current inflation concerns, specifically focusing on energy and food prices.
About Mary Daly
Mary Daly, President of the Federal Reserve Bank of San Francisco, said during several appearances in April through June 2026 that while there is "tremendous investment" in artificial intelligence by businesses, widespread productivity gains from AI are not yet visible in economic data. She described the next year as "the big test" for whether those gains will materialize, adding that firms are still in the early stages of learning the technology and changing their business processes. Daly drew a comparison to the adoption of electrification, noting that sustained productivity gains historically required business process change rather than simply adding new technology to existing operations. She said she is not seeing evidence of financial stability concerns from rising markets or from data center financing, stating that companies are investing "a lot of their own resources" into those projects. On monetary policy, Daly said policy is currently "in a good place" and that the right decision from the May 2026 Federal Open Market Committee meeting was to hold the rate steady. She said providing more specific forward guidance about future rate moves "could be misguided" because of economic uncertainty, and she emphasized the need to balance the risks of overreacting and underreacting to incoming data. Daly stated that inflation remains her "number one priority," pointing to elevated energy and food prices as key drivers, and she said she does not see confusion among the public about the Fed's commitment to price stability. She also expressed that she looks forward to the "real discussions" that incoming FOMC Chair Kevin Warsh has said he wants to have.
Transcript (26 segments)
✨ AI-enhanced transcript with speaker attribution
N
Narrator0:00
San Francisco Fed President Mary Daly admits she hasn't seen any real productivity gains from AI yet, despite the trillions in hyperscaler capex, but remains bullish and says next year will be the big test, conveniently buying time for the Fed while the market crashes on hot jobs data and rate fears.
M
Mary Daly0:16
Interest in it last year and now I am seeing tremendous investment in it. Thinking about how do they train their workforces to be AI ready? How do they think about what AI can do not just in the back office but in the front of house operations? And we're seeing this in small businesses, medium and large, in global companies and more regional ones, and importantly in everything from agriculture to machining and building things to services. And I think that's really the place we haven't seen widespread productivity gains yet. The ROI is still to be developed, but I'm definitely seeing the enthusiasm, and it's picked up tremendously in the last year.
I
Interviewer0:54
We haven't seen the productivity gains yet. You're being very clear on that. As you remember, I asked you over and over again for an hour, 'Please show me the productivity gains.' But in the economic data, whichever set of data you want to look at, do you see an impact from AI, negative or positive?
M
Mary Daly1:14
It's really hard to take that we have had productivity growth that's been outside of the historical norm, and I think that's a positive for the US economy. Everyone wants to say that's AI. What I think of it is: sure, it's possible that businesses are looking for cost savings and they hire fewer workers and they do just as much because they're using an LLM assistant to help, but we just haven't heard from businesses that they're seeing transformative ongoing productivity gains yet. And they want to always underscore 'yet'. And so then I said, 'Well, what's the time frame?' And they said, 'Next year, year after,' because what we know is it isn't just about getting a model and using it for things or an agent, it's about transforming your business processes so that you really take advantage of things we don't even think about today. What can be done differently that will transform the economy? So you can definitely find a single business or sectors who are using it and seeing the gains, but we haven't seen that across the economy going forward. But I'm pretty bullish. I see the possibilities, and I'm hearing more and more that people are seeing early rewards and really recognizing that next year is the litmus test.
I
Interviewer2:24
That bullish is an interesting turn of phrase. Today, putting aside the slight dip in the market, the exuberance that we have seen in financial markets to want to back these companies. And we're about to get more public companies coming, and more liquidity, more money. Is that in and of itself a financial stability issue? Are you worried about the markets riding so high?
M
Mary Daly2:47
You remember, who's doing most of this investment, the Mag 7, who's really there. This is actually something they can do and fund, and their enthusiasm is real. They see what's possible, but I don't think that we should think, 'Oh, there's a financial stability concern' just because the market's gone up. It could go up or down as it has in the past, but a financial stability issue would mean it spreads to the banks, it spreads to consumers or businesses. Right now, I'm not seeing evidence of that. We keep our eye on it for absolute sure, but what I am seeing is that companies other than the technology enthusiasts, companies outside of technology, are using AI to think about how to do their business better in real ways. I mean, I was just meeting with some machine manufacturer; they make machines for a living. That's what they do. And they're thinking about, 'How do I scan in 50 years of plans of these machines I built for companies, and then use those plans in a model to generate innovative new ideas of things I can sell that will be faster, better, cheaper than things I've sold before.' We toured a robotics company that builds things that help manufacturers do better in terms of shipping and distribution. These are real things that do have a capacity to change the economy. So, that's why that's the underlying part of my bullish. It has less to do about the investments that tech companies are making and more about the investments that everyday regular companies that make things and provide services, the things they're doing.
I
Interviewer4:17
President Daly, inflation is still the biggest risk. And those are not my words; we heard it all week long at the Bloomberg Credit Forum, for example. One of the things that you and I have discussed in the past is the massive capex commitment, build-out of data centers in conjunction with a bottleneck in some core areas like memory — is that inflationary or is it disinflationary? That's the thesis that a utility like PG&E would argue: disinflationary because the big guys are buying in aggregate. I still don't understand where we are with that.
M
Mary Daly4:53
It's a timing issue in my judgment. In the beginning, of course, when companies want to invest in big construction projects, a lot of electricity demand, then the companies that are providing those things or areas that are providing those things are going to see competition for the limited amount of services they have. But what they're building creates the infrastructure, the data centers create the infrastructure, or if the big guys come in, as you just said, and help with electrical plants that help with electricity generation, that eventually can help with the prices of those things.
I
Interviewer5:27
Here and now.
M
Mary Daly5:28
But you just have to think about the timing. And one of the things that is really important in policy making is that we not assume we know. We actually look for the evidence. So, what are we seeing in prices today? What are we seeing in the forecast of prices tomorrow? And then how do we think about policy? So right now I'm focused on other energy prices, oil prices, and food prices are driving up inflation. What we do know is that down the road these things could maybe compete for services and costs and raise costs, but we haven't seen real evidence that that's the limiting factor. The limiting factor seems it's hard to get generators, it's hard to get the infrastructure equipment you need, and so you see the big tech companies thinking about solutions they can provide for themselves.
I
Interviewer6:10
The inflationary data must be fascinating here in San Francisco. When I think about all these companies potentially going public, what that means to the employee base, what that means to your house price, and the ability to be able to buy yet more real estate here, what it means for the cost of labor as well. Is there an inflationary pressure that you're seeing here in San Francisco?
M
Mary Daly6:27
I've been through — I moved to San Francisco in 1996, and then we had the dot com, and I know what it feels like to not be able to rent a place that's affordable because the people who were making money were orders of magnitude more. And I was in a position where that's manageable, but that's what's happening: people feel like they're getting crowded out because other things are happening. But that's more about the supply of housing than it is about the demand for housing. We want people to come and invest in this community. We want the city to thrive. We want regional activity to thrive and employment to grow. But as you said, with the more interest people have in living in a place, if you have limited supply of housing, then you're going to have a run-up. So, those are the things that not the Fed, but other policymakers, other federal policymakers and local policymakers like the mayor are working on. But it looks very — you'd see the elements of 1996 already here where there's productivity growth, there's enthusiasm, etc. But the data is not...
I
Interviewer7:25
Is there a risk that we're in that same?
M
Mary Daly7:27
The dot com was very different than the AI boom. And so I just want to — there's a lot that's already being put into businesses, and it's very pervasive. It's not just the dot com. So, I jump to the conclusion that if it has similarities to the '90s, it's going to be the '90s.
I
Interviewer7:44
Well, if you made a presentation, you went out and did some of the most important work in that era of how the advent of the internet would change the economy. So you've kind of established where we sit right now. Inflation continues to rise. How likely or unlikely would that make a rate cut in 2026? How do you tie the two together from this juncture onward?
M
Mary Daly8:06
You know what? I think one of the questions I get asked regularly is what's the path for the rates going forward? And the answer I give, because this is how I really think we have to think about it, is we don't know how the economy's going to play out. We have, as we've been talking about, this tremendous possibility with AI, but we have at the same time the war in Iran that is with an uncertain end, which has pushed oil prices up and fertilizer prices, which have filtered into food prices. And right now, those are fairly contained, and if you look at the futures market for oil, it's $80 a barrel by the end of the year. But we have to think about that. And so, right now policy's in a good place. We are prepared to respond either way, whatever the economy brings, but I think giving more forward guidance about what's possible could be misguided in the end, because we just have to wait for the economy to evolve. Everybody wants to resolve the uncertainty today, but I think that's a mistake, because it will close off our mind about what we really have to look at: the inflation risk possibility, but also the possibility the war ends, oil prices come back down, and we're back to the underlying dynamics with some of the positives of AI we've been talking about.
I
Interviewer9:19
Can I ask you about the labor market as well, because you tried to bring this transparency of how you're thinking about things with your blog, for example. And one of the really interesting ones that I was reading and catching my attention was the idea that we've got zero labor growth now. Immigration's changed, the way demographics changed. Are you feeling like we've seen some improvement, some resilience in the labor market? And will that hold? Has it firmed?
M
Mary Daly9:43
I think that's too early to say firmed. I think we're — there's always statistical errors. You can bounce around from month to month. But I was one of the policymakers who were a little worried about the labor market at the end of last year. I was very supportive of the cuts that we took to stabilize conditions there. So relative to that point, I think we've really stabilized, and I'm starting to see businesses feel a little more cautiously optimistic, which will feed through to hiring. But they're not just running out to hire people. They are being cautious. They can get an agent. They're interrogating how much AI can do for them before they hire. And regularly we talk to our businesses and they say, 'We don't want to hire a bunch of people, find out AI can do certain things and we need a different set of skills. So we want to wait. We want to be patient on our hiring and make sure we're not over hiring because if you ever go through a period of time where a business has to lay workers off, it's a painful experience for the workers and for them.' And so they just don't want to get overly confident only to find out they have to make adjustments. So that caution will be with us for a bit.
I
Interviewer10:51
We are live on Bloomberg Television and Bloomberg Radio. We are in San Francisco and we're at the Bloomberg Tech event, and we're speaking with San Francisco Fed President Mary Daly. Reset a little bit, but if I may, it's the first opportunity we've had to ask you. Have you spoken to Chairman Walsh about how he sees the Fed evolving? About changes to the institution. And if you may, fold in the context of your district, San Francisco Fed — yes, much more than that — and your role going forward.
M
Mary Daly11:22
Sure, absolutely. So, I think what we really want when any new chair comes in, and what we want from all of our leaders of Federal Reserve Banks and all of our governors, is that you're thinking constantly about how can the Fed be better? How can it better serve the American people? How can we do our work more efficiently, more effectively, and more resiliently? Ultimately, everything we talk about is put between two bookends. We are fiduciary stewards of public trust, which means we better have services that people can depend on, and we better work hard to achieve our goals. And we're fiduciary stewards of public funds, which means we are very careful about how we spend taxpayer dollars. So with those two things in mind, I mentioned I joined the Fed back in the '90s. I joined the Fed in the '90s, and we had check processing. Remember checks? We had check processing. Everywhere we had a location, we had people who processed checks. But then check demand started to fall, and so we consolidated those activities into a few locations. And that level of modernizing, constantly thinking about how can you do better, is what I see now.
I
Interviewer12:31
Is the chair's equivalent of that present day?
M
Mary Daly12:33
He just joined. So, I'm going to give him the right time he has to announce that. He's talked about making sure he's holding on to that tradition, and he comes in with a lot of ideas. But what I've heard him say again and again, which I really appreciate because he's the fifth chair I've worked with, all of the chairs I've worked with have the same basic compass. It is to do our best work for the American people and work with all the individuals who are earnestly doing their work in the Fed to do it well. And I see that in Chair Walsh.
I
Interviewer13:06
That earnest work within the Fed, how much of that is being modernized to adopt AI? How hard or easy is it at this moment when you're such a regulated institution in and of itself?
M
Mary Daly13:15
Well, we're careful, like all regulated institutions and importantly like all businesses. Businesses, when I talk to businesses, the last thing they want to do is take huge risks that destroy their shareholders' value, the value of their company. The same is true for us. We're good fiduciaries, stewards of public trust, and good fiduciary stewards of public funds, which means we're always driving to adopt new technology, to do our work more efficiently, but we recognize we have to do that safely. People want to know they can get their money when they need it. They want to know that the banks are well supervised. They want to know that monetary policy is not made by machines, it's made by people making judgments not only about models and rules, but also about the lived experiences of people across our country.
I
Interviewer13:57
Very quickly before the show ends, what are you seeing in credit? That's a big story for us in how data centers are financed. For some, it's very worrying.
M
Mary Daly14:08
Well, we're watching that carefully. I watch that carefully. What we do see is that there's a lot of those companies investing a lot of their own resources in those. So, it's something to keep our eye on, but at this point, again, if you...