AI and the global economy: Productivity, jobs, and inflation

The Angle Music

Cold open: “So, the list of shocks is very long, and some of them will be positive and some will be negative. So, this may sound messy, and also confusing, but I think it's the nature of living through a period of major technological disruption.”

Jennifer Martin

Welcome to The Angle from T. Rowe Price, a podcast for curious investors. I'm your host, Jennifer Martin, a global equity portfolio specialist at T. Price Associates here in Baltimore, Maryland. This season, we're exploring the rapidly evolving artificial intelligence landscape and what the future might hold for investors and innovators alike. In today's episode, we're diving into how the AI revolution is impacting the fixed income markets, which have become a key funding source for AI CapEx and the long-term implications for productivity, inflation, and interest rates.

Today, I'm joined by two T. Rowe Price fixed income experts, Arif Husain, head of global fixed income, and Blerina Uruci, chief U.S. economist. Thank you both for joining us today. Blerina, you and I recorded a podcast together on AI back in 2024, which feels like forever ago. What has changed in the macro implications of AI in the last two years?

Blerina Uruci

Thank you, Jen, it's great to be here today. And yes, I went back and listened to our podcast from two years ago. It was fun. We had fun. It sounded like we had a good session. So back then the discussion was very theoretical or hypothetical, if you like it. We're talking a lot about hallucinations. When are we going to get better models? Adoption, speed a little bit, people hesitating to adopt, and so on. And I think right now the discussion has really shifted towards show me the numbers. In all the discussions that I'm having with our investment staff as well as our clients, I'm having to talk about how AI is affecting growth, labor market, and inflation, and what's the outlook.

So let me provide some statistics around this. For example, in the last six quarters, if you look at AI related CapEx and US GDP, about half of US GDP growth has been driven by AI CapEx. Now, this is very significant because the three years prior, this contribution was a third of what it is today, so really we've seen a pretty big impact on growth numbers themselves. And although this AI models may live in the cloud, to me it feels like the investment cycle is really very physical. It's showing in the GDP data.

Labor is very interesting and it's much more nuanced. Whenever we survey US workers about the effect of AI, they report being worried that it will replace them. But I think this is also impacting wage pressures in a way, because it makes workers more reluctant to ask for bigger wage increases. Then when we look at layoff numbers per se, we're not seeing there, what workers are worried about, but we are seeing slower net job creation. And for me, this is important, and it brings me to the last point in terms of macro data. To produce output for the economy, we need labor inputs, we need capital, and then we need this magical unobserved ingredient that is called productivity. Now productivity as we measured it, output per worker, has accelerated significantly from 1.5% annual growth before the pandemic to 2.5% in recent quarters. So, after this productivity trend stagnating for two decades, we're finally talking about a productivity revival.

But AI per se is a global story. And it's a very important story for financial markets globally. Let me tell you why. So, as you know, we have a fantastic fixed income and fantastic equity fundamentals research here at T. Rowe Price. Our analysts have gone and crunched the numbers, looked bottoms up, how much CapEx, AI CapEx spend we're going to see in 2027. That's over $1 trillion. I love putting this statistics into perspective. That's 3% of US GDP. But now, in real time, we're seeing fantastic speed in CapEx expansion and how that affects the economy. And the next point is that when you ask our analysts, they will tell you two thirds of that spend is going to be on imported chips and overall tech goods that other countries are producing. So, essentially what we're going to see is a 2% of GDP stimulus from the US private sector to the global economy. But I think it's a broader global story about the US CapEx cycle, making the global economy more resilient. And this is going to have implications both for fixed income and issuance as well as equity markets globally.

So, to go back to your original question, what's interesting to me is that the vocabulary around the AI discussion has changed a lot, but the framework has not. We're still talking about CapEx cycles. We're still talking about productivity, and labor, and inflation.

Jennifer Martin

That's perfect. So what I heard was, big impact on growth, both physical and digital, productivity increasing, CapEx probably still expanding. You gave a trillion, I think it may be higher after some of the reports coming out over the last few weeks. And US exporting growth globally, so accelerating GDP, that's really a nice way to start. So, can we distinguish between near-term and medium-term effects on the macro economy? If we think back to widespread adoption, for example, of the internet, there were immediate implications, as well as longer run changes that shaped how the global economy works today. And I remember the last time we talked about your vacation planning with ChatGPT. I was just thinking about that.

Blerina Uruci

I may have, yes, I had another personal anecdote to bring to that, but we can talk about it later. Anyway, so we can talk, of course, about the near term and long term effects of AI and this CapEx cycle. And I also feel like this is the prevailing market narrative right now. That's what everybody loves to talk about. So let me get into that discussion, and also why I think it's the wrong way to think about innovation generally and the AI cycle more specifically.

So, what is the consensus view? AI is going to be inflationary in the near term because it's resource intensive, creates price inelastic demand for capital. This hyperscalers, they'll pay any price, they can afford to pay any price for the goods that they need. And then it's going to be disinflationary in the long run because it expands the supply side of the economy. And let's look at how good productivity growth was during the 1990s. That's a great example, right?

So, I see the appeal of this kind of thinking. It's clean. Everything fits in its own box, in its own time horizon. You have some frictions in the near term, but then in the long run or medium run, everything is fine. But I think a better way, even if it's messier, a better way to think about AI is a series of shocks, in my view. Like we have a CapEx shock right now. We can have an adoption shock, a price shock, a labor demand shock. And let's not forget, there will be a shock to the production structure of the economy where we have well-established oligopolies. Are they going to become weaker or more powerful as a result of AI?

So, the list of shocks is very long, and some of them will be positive and some will be negative. So, this may sound messy, going back to that earlier point, and also confusing, but I think it's the nature of living through a period of major technological disruption. And then we will look at history to inform our decision making. But at the end of the day, we also recognize that AI is a very unique shock.

Jennifer Martin

Arif, let's maybe have you add on to Blerina’s points of thinking about near-term and medium term effects on the macro economy.

Arif Husain

Yeah. Well, if I'd rather just listen to Blerina’s holiday plans, to be honest. But if you're insisting on bringing me in, I was loving listening to you. I think, let me say a couple of things. So timeframe. Time horizon is generally one of the most important but most ignored investment factors, whatever we're talking about. And, you know, the one thing I've learned through my career is nothing ever goes in a straight line. And it's never as obvious as the consensus. But broadly speaking, I do think that differentiation between the short, medium, and long term that Blerina brought out is, is broadly, stylistically the right one.

You know, if I think as far as I can see right now, I do see a couple of things. So, firstly, in the very, very short term, we haven't just got AI, but we're also facing a lot of other constraints. We've got compute constraints, we've got supply constraints, and we've also got energy constraints because of what's going on in the Middle East at the moment. As I think a little bit longer term, I also wonder about how AI adoption, how governmental influence will will feed in. You know, let's think about it. You know, if you follow that logic through, we're going to end up with a lot of, a high degree of youth unemployment. You know, I'll throw it back to you. Has there ever been a good outcome of a high level of youth unemployment in the world? It's just, so I do think there's going to be a little bit more intervention, and it just won’t be as linear as everyone expects.

Short term, for sure, I do think this is inflationary. My view, and I'm not an academic like the Fed, I'm as an investor, I can't rely on this mythical beast called productivity, riding over the hill and and solving all the problems. Just just can't, right? So I am somewhat skeptical there. I also think, you know, Blerina is really correct to say, you know, the US is at the very front of this. But I also think there's a geographical element to think about as well.

If if you're a, you know, if you're a country with a very mature labor force, where you, you're, call it demographically challenged, and that's been your problem for a good while, then AI is so very welcome. It's going to revolutionize your economy. However, if you are a very young society where you know your demographics are skewed towards a much younger cohort, the way it's going to be implemented, the way it's going to impact is going to be very, very difficult. So, you know, just just top of my head, I think Japan versus India, for example, the implications through developed markets and in emerging markets can be very, very different.

Jennifer Martin

Well, I think both of you brought up some excellent points talking about. I love that idea of series of shocks, because it's true. And I also think things happening not linearly is also very real. And, you know, Arif you brought up some excellent points related to regulatory risk, the inflationary risks that many people are monitoring. And most importantly, I think you introduced the new topic of demographics that will bring different challenges depending on where you're domiciled. So, I guess that brings us to maybe what might be the most significant question everyone is currently asking on how AI will affect inflation and labor markets in the longer run. Arif, you introduced some of that, but maybe Blerina, you could add on to that.

Blerina Uruci

Yeah, I think I wanted to go back to Arif’s point about near-term inflationary effects and so on, because it's very important. And the near term matters, it’s what we're going to see first,  right? But then I also on inflation, I want to go back to the discussion about time horizons versus multiple shocks, because I think the market consensus right now perfectly illustrates the point that I was trying to make.

So, what's the consensus? In the near term, demand for tech goods will increase. And that will, because we have finite supply of these goods that will increase prices. And then we can go and find the CPI components where this shows up in terms of tech goods. We can look at electricity demand because of data centers. And then the consensus is, it's gravitating towards these 2/10 or 3/10ths increase to inflation every year. So annual inflation every year will be 2/10th higher because of this AI demand over the next few years. We don't know exactly what what the time horizon is. But at the same time, we've been discussing the effects on labor and on youth unemployment, how workers are feeling nervous about AI in the workplace, and then how that might affect wage demands.

So then how are we measuring the fear effect of AI on wages? How are we measuring the fact that we have labor supply very limited post the immigration policies of this administration, but not a tight labor market? And we have strong productivity growth. I think these are important factors to keep in mind and highlight this idea that it's not simple to find the net effect on inflation. And that's why we can't just think near-term is a demand shock and long term it's a supply shock. It's a lot of shocks happening at different times. And we can't know for sure exactly when.

So then, what is my mental framework for this? I like to think of the shocks that will affect labor and inflation one by one, and then I'll try my best to determine the sign because of this shock, inflation up or down, labor up or down? And also remember to be humble about false precision. We can't know for sure the magnitude. If we can get the direction right, and the net effect of these shocks right, I think we're going to be doing great.

And then finally, I wanted to share an anecdote here from a recent investor trip that I did. We were visiting data centers and distribution centers, and distribution centers that were adopting a lot of automation on the floor of the warehouses. And I had a conversation with one executive that really stuck with me. He was talking a lot about how they had not fired any workers over the last five years. Employment numbers were flat. And pushing against this narrative that robotics will replace jobs. He was saying, look at my payroll numbers. They haven't changed. So then I asked, what has happened to revenues and profitability for you? And it had increased many folds, but the wage bill had been stable. To me, this is real time, real life productivity shock.

Jennifer Martin

That’s a good example.

Blerina Uruci

Yes, because we are not; we're thinking about firing, but net job creation when output is expanding at a fast pace, I think that's a shock of its own. And it's telling you where the profits are going in the economy. And so, I guess what I wanted to bring to this discussion, on top of what the points we've made so far, is that as macroeconomists, we look at aggregate labor and that can look fine on the surface when things under the rug for young professionals are not as strong as they look for the bottom line. And that goes back to your earlier point Arif. So, this idea of who's being affected, how are young workers being affected, and how are profits being distributed will matter a lot when we think about the net effect of AI.

Jennifer Martin

And Arif, do you want to add some insights about how AI will affect inflation and labor markets in the long run as well?

Arif Husain

Yeah. So, just picking up on what Blerina said, I think, just she said some amazing things that just make a lot of sense. But the thing that really resonated with me was, it's just not going to be as easy as the consensus thinks about. And, you know, I think about the, the, the narrative from Fed Chair Kevin Warsh. Throughout his confirmation hearings, as he's become the Fed chair, I, I've heard a narrative from him that productivity from AI is going to keep rates down. It'll allow them to ease central bank policy. And just reflecting on what Blerina said, I just don't think it will be that easy.

Jennifer Martin

And as we know, every Fed governor, at least every new Fed governor, gets tested, so well, we'll see. We'll see what test he gets. It's going to be a pop quiz maybe.

Arif Husain

I will I will say, Jennifer, that Mr. Market is is a bully. And in classic bullying behavior, it likes to haze the new kid. And, and so, we have a new chair, let's see what the market does.

Blerina Uruci

And also just a reminder that we're recording this podcast the day after the July FOMC meeting, where Mr. Market was really pushing the Fed narrative, especially in the long end of the curve.

Jennifer Martin

Well, thinking about the Fed and interest rates, do you have a framework for analyzing the effects of AI and interest rates? For example, how will this pan out in terms of the impact on rates going forward?

Arif Husain

Yeah, we've talked a little bit about constraints, and I think what a lot of people fail to understand is the is the financing constraint. So, you know, to build out the capacity needed to meet this, this dream, is is a ton of CapEx. And the way that is being financed primarily is through debt issuance. And, you know, we're seeing that debt coming from every corner. We're seeing it in an unsecured debt issuance from the hyperscalers. We're seeing it from unsecured on data centers, and we're seeing it through the investment grade market, the high yield market, the securitized market. Wherever you look, you are getting AI related debt. And I think the most important thing to say is that debt is competing for shelf space. It's, it's, it's joining an arms race that already exists between every other debt issue out there.

Think about the amount of post-Covid deficits that governments around the world are running. They need to be financed through debt issuance. And so effectively, we have a competition for capital. And the AI issuance is just joining in. So, I think the natural conclusion, and actually we're already starting to see this, the, the spreads, the credit spreads on the long end debt of of the hyperscalers are already starting to widen out. The cost of debt for them is starting to widen out. I think we're starting to see that pressure feed into US Treasury yields as well. Ultimately, I see all of this leading to higher financing costs in the markets. And that that that matters, right? That that again is a constraint. Because it's not just for the, the hyperscalers and for AI, it's for when you guys walk in and into your local bank and want to get a mortgage.

Jennifer Martin

As you were talking, it made me think that, you know, the financing costs are increasing. And what's notable is you're not seeing any real credit drivers. You know, fundamentals of many of these hyperscalers, for example, are accelerating. So you're you're suggesting we're going to have this competition for capital independent of that right now.

Arif Husain

It's you know, think back to economics 101. You know, the two lines demand and supply. And that's ultimately what we have. The price of debt needs to increase. The yield needs to, you know, the needs to increase, because we have so many borrowers who need to come and take debt from the market.

Blerina Uruci

And I think Arif since you said Economics 101, it made me think about Finance 101 and how you would approach this nominal yield problem, and you try to separate within that overall yield, the Fed path, the inflation risk premia, the term premia. And we can really see how AI affects all of those because it affects growth, it affects the inflation path. Perhaps it's less obvious how it might affect risk premia or term premia, but I'll bring one of your favorite topics here, which is bond supply and fiscal deficits. Because everybody assumes that because we're going to have so much growth, deficits will go down because we'll raise tax revenue. But I think that really depends on whether those profits will go to workers because we tax income at a higher rate than capital, or whether those profits will go to capital and corporates. And I think that has the potential to increase term premia.

And then, Jennifer, to your point, I was thinking of a very interesting paradox because we think long run, at least everybody seems to agree that AI is going to lower inflation, therefore the Fed can cut interest rates. Right? And inflation interest rates have to be lower. But I think what people miss is that you can have lower inflation and higher interest rates because of the implications of AI on R star. So, AI can increase potential growth, which can increase R star, and essentially means the economy needs, because there is so much demand for investment and capital, the economy needs to run at a higher equilibrium interest rate. So, I feel like everything is pointing toward this world with higher interest rates for longer.

Jennifer Martin

No, I think it's a great insight. And I think Dave Eiswert had mentioned that on another episode where we would have, maybe, an interesting scenario of higher GDP growth but higher inflation. And that's kind of a new paradigm to also invest behind.

Blerina Uruci

That's also very interesting angle.

Arif Husain

Yeah. Maybe just picking up on that. I mean I’m going to bring out my fixed income bear persona right now, because I think just trying to get a little bit of balance. So, you know, this idea of not going in a straight line, this being complex, shocks following on shocks. I think, you know, we do need to think about, certainly, as we talked about, the supply of, of, debt. You know, one thing I know as a bond investor is, I don't like negative cash flow. And, if I extrapolate that point, I can imagine there being a point where something goes wrong. Again, I don't want to over-index to the negativity, but, you know, everything to do with AI is this, is this very positive dream. I don't think we get there without some sort of stumble.

And given how important the CapEx, given how important this is to businesses and the macro economy, I think any stumble in AI could be become systematic. And so there is a, actually, a very, I think, plausible scenario where we're with the Fed having to react to a negative growth shock because of a systematic event within AI that creates a need for cutting rates. But, you know, just wanting to throw out, pop the bubble of of potential optimism there because, you know, as Blerina said, there's a lot of scenarios on the table.

Jennifer Martin

I love the balance. Now I think it's good to be balanced on that. And so, in an earlier episode of this season of The Angle, Dom Rizzo mentioned that he thinks AI will eventually create jobs but is likely to worsen inequality, which is something we've kind of hit on a little bit today. So maybe, what's your view on the distributional effects of AI? How could that impact fiscal policy in the future? And I think both of you have kind of touched on it. But let's, let's, bring it home.

Blerina Uruci

That's true. We've touched about it throughout the the other answers. So, here's how I would think about this question and this problem more broadly. We can think about within the economy distribution effects, and we can think about the global distribution effects. So, let's take the first problem first and focus on the US again. Because AI here is creating a lot of profit. And it will continue to do so for the years to come. And the question is, who will capture those profits? Will it be capital or workers? Will it be high-skilled workers, lower-skilled workers? We can use our imagination here, but we can also look at history, right? And in macroeconomics we have this concept of labor share, which is the amount of income produced in the economy that goes to labor, and the profit share, which is the amount that goes to capital or companies.

So, what we see after every big productivity spurt is that businesses, they're very good at capturing the profits. And so the business share has continued to increase. And the labor share has been on trend decline for many years. And when you look at this decline in the labor share, I think it goes hand in hand with a cost-of-living challenges that many consumers are reporting and facing, and the rising inequality in modern societies.

Now, the other interesting angle we can bring to this. So, we have labor share and capital share. Let's think about the fiscal implications of this. Because in the current tax system, not just in the US but globally, we tax income from labor more highly than income from corporates or from profit. So, if we keep the tax system the same, and all, most of the profits go to the profit share to capital, then revenues will not increase, fiscal revenues for the government will not increase as much as people expect. That's one point to keep in mind.

The other point is governments may need to spend more on those workers that are being replaced or displaced by AI, on training, on income support, and so on. So that's pointing towards this fiscal situation where larger deficits may be with us for a bit longer if we don't change anything in terms of how we spend money, spending priorities, or tax priorities. So, this goes back to your point about deficits, Arif, and then whatever we discussed about the US or within economy distributions, we can apply that to the rest of the world. And then we can add another layer here that countries that are at the frontier of AI, like the US and China, are going to do better because they're going to sell this product to the rest of the world.

Jennifer Martin

Anything to add, Arif, to that?

Arif Husain

Honestly, this, I think we could open up a real can of worms here, but I think the thing that really struck me as I listened to Blerina is, you know, right now the political system is set up to tax individuals and businesses differently. I don't think it's a controversial point to say we could easily see a swing in politics as wealth inequality, or the, those, profits flow into different places. Businesses don't get to vote. The people do.

Blerina Uruci

That's an interesting point because it talks to the rise of populism, and we're living in that way right now. And.

Arif Husain
Exactly.

Blerina Uruci

Whether that continues, and what are the implications for for our current economic system? It's a, it's a pretty big topic.

Arif Husain

And that's not just a US thing, right? It's it's.

Blerina Uruci

Absolutely.

Arif Husain

It's, it's, happening around the world in different microcosms in different ways.

Jennifer Martin

Wow. That's a, that's a heavy one to end on right now. But I, but I do think it brings your imagination. I think everyone can relate to, depending on where they're listening to this, they've all experienced a little bit of element that in their local and probably more national politics. So. I, I appreciate that perspective very much.

Well, you know, I would just say that might be a perfect place to end on. Very good insights from both of you. And on behalf of our audience, thank you for your participation today. We really appreciate learning from two experts in the field of fixed income.

Blerina Uruci

Thank you both. This has been fun.

Arif Husain

Thanks for having us.

Jennifer Martin

Again, I'm Jennifer Martin, thank you for listening to The Angle. We look forward to your company on future episodes. You can find more information about this and other topics on our website. Please rate and subscribe wherever you get your podcasts. The Angle. Better questions, better insights. Only from T. Rowe Price.

DISCLOSURE

This podcast episode was recorded in July of 2026 and is for general information and educational purposes only. Outside of the United States and Australia, it is for investment professional use only. It is not intended to be used by persons and jurisdictions which prohibit or restrict distribution of the material herein.

This podcast does not give advice or recommendation of any nature, or constitute an offer or solicitation to buy or sell any security in any jurisdiction. Prospective investors should seek independent legal, financial and tax advice before making any investment decision. Past performance is not a guarantee or a reliable indicator of future performance. All investments are subject to risk, including the possible loss of principal.

The views contained herein, including forecasts and forward-looking statements, are those of the speakers, as of the date of the recording, and are subject to change without notice. These views may differ from those of other T. Rowe Price associates and/or affiliates.

Economic estimates and forward-looking statements are subject to numerous assumptions, risks, and uncertainties. Actual outcomes could differ materially from those anticipated in estimates and forward‑looking statements. Forecasts are based on Subjective estimates about market environments that may never occur.

Information is from sources deemed reliable but not guaranteed or verified. Please visit http://www.troweprice.com/theanglepodcast for full global issuer disclosures.

This podcast is copyright by T. Rowe Price 2026.

CON01724700

202608 - 5786847