When President Donald Trump chose him to lead the Federal Reserve, Kevin Warsh argued that the artificial intelligence boom would help tame inflation.
More than a dozen of his colleagues, including two other Trump appointees, now say it’s doing the opposite, playing a role in fueling inflation, which could soon mean even higher interest rates on credit cards, car loans and mortgages.
Warsh will get a chance to revisit that argument on Monday, when he opens a conference in San Francisco meant to inform the work of a task force he created to study AI’s effects on jobs and productivity. He heads into it looking increasingly out of step with many other Fed officials.
This week, Fed governor Michelle Bowman, a Trump appointee who like Warsh once predicted AI would help push inflation down, conceded that is not yet happening. She said the AI buildout accounts for roughly a quarter of the excess inflation the Fed is fighting, about as much as the pressure from tariffs or energy prices.
Fed governor Christopher Waller, who said a big interest rate hike should be on the table in December, on Tuesday pointed to research by Fed economists on AI and high energy prices. Their findings, the Trump appointee said, “do not encourage me to think that these factors will be fading anytime soon.”
Hotter-than-expected inflation numbers out this week underscored the concerns raised by Bowman and Waller. Electricity prices jumped in October, and smartphones also logged their largest monthly increase since the government began tracking them separately in 2019. The data center buildout has been driving up electricity prices over the past year, as well as demand for chips that are also critical for consumer electronics.
Warsh has acknowledged that the artificial intelligence boom is increasing the price of computer chips, but he has questioned whether it is driving inflation more broadly.
“Economic growth is higher, but that doesn’t tell me that economic growth is somehow causing inflation and making our job harder,” he said at his news conference last month after the Fed held rates steady a week before the midterm elections. Warsh said politics played no role in the decision.
The debate over AI’s role in inflation comes down to timing. Right now companies racing to build AI are competing for the same chips, electricity and electricians, which pushes prices up. But AI could eventually make businesses more efficient, allowing them to produce more for less and helping hold prices down.
If that payoff comes in the near term, as Warsh suggested before he took over the Federal Reserve, the central bank can afford to be patient. If the productivity gains are years away, as many at the Fed now fear, Fed officials may need to keep raising interest rates to cool things off.
So far, Warsh’s views haven’t kept the Fed from acting. The Federal Open Market Committee voted unanimously to raise rates in September for the first time in more than three years. Several officials have said they expect another increase at the Fed’s December meeting.
Warsh was openly enthusiastic about AI before he took the job, arguing last year in an op-ed that it could help bring about a “new golden age” for the economy if the Fed changed some of its policies.
“AI will be a significant disinflationary force, increasing productivity and bolstering American competitiveness,” he wrote.
At a Senate hearing in July, Warsh called a disagreement over whether AI investments were driving inflation one of the Fed committee’s “good family fights” and said his view “might be somewhat nuanced” from “some of my colleagues.”
“I don’t view a one-time change in prices as necessarily being inflationary,” Warsh told the senators, because he expects higher prices to prompt companies to produce more. He predicted AI would increase “measured prices” over the next year. “But whether that’s inflationary or not, that’s up to the Federal Reserve, and we’re going to have something to say about that,” he said.
Michael Strain, an economist at the right-leaning American Enterprise Institute, said he sees AI’s effect on the economy playing out in two steps. First, companies spend heavily on data centers and other AI investments, and “all of that pushes up the prices.” Then “businesses figure out how to use this technology to make money. That translates into productivity increases, and that is disinflationary,” he said. The challenge for the economy is that he expects that second step “no earlier than 2028 and maybe later.”
Jacob Robbins, an economist at the University of Illinois at Chicago and a nonresident scholar at the Washington Center for Equitable Growth, a left-leaning think tank, said Warsh’s view has sometimes seemed too relaxed about rising prices.
“He has been very persistent with his view of AI, and he’s essentially stuck to it,” Robbins said. He added that Warsh’s embrace of the idea of near-term productivity gains from AI looks rosy next to the data.
AI was barely a topic at the Fed’s policy meetings in early 2025. When it did come up, officials mostly talked about it as a force that could help hold prices down.
But as early as this past February, Fed Vice Chair Philip Jefferson warned that a surge in AI-related spending “could raise inflation temporarily.”
At Warsh’s first meeting as chair in June, most officials said strong AI investment was helping push the economy past what it could sustain, which “could contribute to more persistent inflationary pressures,” according to the minutes.
By July, some Fed officials were pointing to higher prices for everyday purchases, including “smartphones, computer equipment, software, and electricity,” the minutes said. Since then, several officials have gone further.
“While the impact of AI on productivity remains unclear, its effect on prices is not,” Anna Paulson, president of the Federal Reserve Bank of Philadelphia, said in a September speech.
John Williams, president of the Federal Reserve Bank of New York, said days later that AI spending’s effect on inflation was becoming harder to ignore. And Fed governor Lisa Cook said in a speech that month that “in the short term, AI appears to be adding inflationary pressures to the economy, postponing inflation’s return to our 2 percent target.”
On Monday, Austan Goolsbee, president of the Chicago Fed, said at an event that the AI buildout is “starting to look more just like old-fashioned demand overheating,” with everyone “out trying to hire an electrician at the same time.” That kind of inflation, he said, isn’t something that “just goes away on its own.”
Bowman said Tuesday that she still expects AI to “raise productivity and put some downward pressure on inflation” over time, and many other Fed officials share that long-run hope. The question is when.
Warsh’s AI task force that will convene Monday is led by Marc Andreessen, a Trump ally and one of the AI industry’s best-known venture capitalists, along with a Microsoft executive and a Stanford economist on leave at the AI company Anthropic.
Warsh has said Monday’s conference in San Francisco is meant for “a lot of listening,” including on whether AI will make it “harder or easier” to keep inflation in check over the next several years.
“Is this a productivity miracle that’s going to happen, or will it happen slowly?” he said at last month’s news conference. “I know I, for one, am keenly interested in refining my own views on that question.”
Federica Cocco contributed to this report.
The post The Fed chief said AI would tame inflation. His colleagues disagree. appeared first on Washington Post.
