Cleveland Federal Reserve President Beth Hammack just put a price tag on the AI gold rush, and it is not denominated in GPU orders. In a CNBC interview on Monday, Hammack warned that “insatiable” demand for AI infrastructure is helping fuel persistent inflation and that the Fed may need to raise interest rates to bring prices back under control.
The warning carries weight. Hammack is a voting member of the Federal Open Market Committee this year, and her comments landed just two weeks after the FOMC held rates steady at its June 17 meeting while penciling in a quarter-point increase before year-end. She is not speculating from the sidelines. She is telling you what she might vote for.
The Paradox Wall Street Does Not Want to Hear
Here is the structural tension Hammack is naming, and it is one the market has been studiously ignoring. The same AI spending wave that has lifted equity valuations across the semiconductor, cloud, and data center supply chain is simultaneously generating the inflationary pressure that could force the Fed to raise rates. Higher rates compress the very valuations that the AI boom inflated.
Speaking with CNBC’s Sara Eisen at the European Central Bank conference in Sintra, Portugal, Hammack cited a manufacturer in her district that builds electric switching equipment for data centers. “The demand is insatiable, that these companies, these hyperscalers, will pay almost any price for those inputs, and they need things built yesterday,” she said.
That is not a theoretical concern. When the largest companies on the planet are bidding up industrial inputs with unlimited budgets and zero price sensitivity, they are not just building data centers. They are building inflation into the cost structure of every industry that competes for the same labor, materials, energy, and construction capacity.
Five Years of Too-High Inflation
Hammack did not sugarcoat the broader picture. “We’ve got inflation that’s too high, and it’s been too high for the past five years,” she told Yahoo Finance. The May CPI reading came in at 4.2%, a three-year high driven by energy costs and sticky services inflation.
Her policy prescription was equally direct: “If inflation continues to persist at these elevated levels and I don’t see any restraint from policy, we may need to raise rates to bring that policy restraint in and to bring inflation back down.”
She also dismissed the idea that current rate levels are slowing the economy in any meaningful way. “When I look broadly, particularly around large companies, I’m not seeing a lot of restraint in the economy. I’m not hearing from these businesses that interest rates or credit spreads are a reason why they’re holding back from investment and growth,” she said.
That observation undercuts one of the core assumptions embedded in current market pricing: that the Fed is already restrictive enough and just needs to wait. Hammack is saying the data does not support that thesis.
The Warsh-Hammack Divide
Hammack’s position sets up a meaningful split inside the Fed. Chairman Kevin Warsh, who ran his first FOMC meeting on June 17, has argued that productivity gains from AI will ultimately prove disinflationary, lowering the cost of labor and goods over time. That view has provided intellectual cover for markets to treat the AI spending surge as growth-positive without an inflation offset.
Hammack is making the opposite case: before those hypothetical productivity gains materialize, the physical buildout itself is generating real, measurable inflation right now. Electric switchgear, copper, concrete, construction labor, electricity, cooling systems, land for data center campuses. The hyperscalers are not waiting for unit economics to pencil out. They are spending as if the race is existential, because for several of them, it is.
The question is not whether Warsh or Hammack is right in the long run. It is which effect dominates in the policy-relevant time horizon of the next 12 to 18 months. If Hammack’s read is correct, the Fed’s penciled-in quarter-point increase may be a floor, not a ceiling.
What This Means for the AI Trade
The market implications are concrete. A Fed that raises rates because of AI-driven inflation would create a perverse feedback loop for the sector. Higher borrowing costs raise the discount rate on future earnings, compressing the valuations of the very companies whose spending is driving the inflation. Microsoft, Alphabet, Amazon, Meta, and Oracle have collectively committed hundreds of billions to data center buildouts through 2027. Those capital plans are not contingent on rate levels, but the equity multiples investors assign to those companies very much are.
Bond markets are already pricing in elevated risk. The 30-year Treasury yield has been hovering above 5% since mid-May, reflecting persistent inflation expectations that predate Hammack’s comments but validate her thesis.
For the broader economy, the AI inflation channel Hammack identifies adds a new variable to an already complex picture. Tariff-driven import cost pressure, energy price volatility, and sticky services inflation were already in the mix. Now add unconstrained corporate capital spending in a single sector as a distinct inflationary force, one the Fed has not previously had to model at this scale.
The Bottom Line
Hammack is saying something the market needs to hear, even if it does not want to: the AI boom has a cost that extends beyond the capital expenditure line on a quarterly earnings report. It is showing up in inflation data, and a voting Fed official is willing to raise rates to address it. The assumption that AI is purely disinflationary because it will eventually boost productivity is a bet on a timeline that the Fed cannot afford to wait out. Prices are high now. Workers and consumers are paying now. And Hammack is making clear that if the data does not turn, her vote will reflect it.