The rapid expansion of AI infrastructure is not easing inflation yet, according to Federal Reserve Chair Jerome Powell. Instead, Powell said the current wave of data center construction is adding pressure to prices across parts of the economy, as utilities, builders, and suppliers race to meet the demands of a fast-growing sector.
Speaking after the Federal Reserve decided to leave interest rates unchanged, Powell said the physical buildout behind artificial intelligence is creating near-term cost pressures. In his view, the economic story is not yet one of lower prices driven by efficiency gains. For now, Jerome Powell data centers inflation is a story about higher demand for land, electricity, equipment, and construction inputs.
AI’s infrastructure boom is creating real-world price pressure
Powell pushed back on the idea that advances in artificial intelligence should already be translating into lower inflation and lower interest rates. He acknowledged that AI could eventually improve productivity in meaningful ways, but said the current phase looks more inflationary than disinflationary.
His reasoning was straightforward. Before businesses and consumers feel the full economic benefits of AI, companies first need enormous amounts of physical infrastructure to support it. That means new data centers, grid upgrades, cooling systems, semiconductors, industrial equipment, and labor, all of which raise demand across multiple sectors at once.
In that environment, prices can move higher rather than lower. Powell said the immediate effect is pressure on the goods and services required to build and operate those facilities. That helps explain why the Fed is not treating AI as a near-term argument for cutting rates.
Where are the productivity gains everyone expected?
The debate around AI has increasingly centered on whether automation and software improvements will lift output enough to lower costs across the broader economy. Fed officials recently raised their long-run growth estimate, a sign that stronger productivity may be taking hold. But Powell made clear that better long-term growth does not automatically mean inflation falls in the short run.
That distinction matters for markets and households alike. Productivity gains can take time to spread through the economy, especially when the early phase of adoption depends on massive capital spending. Powell suggested that, at least for now, the demand created by AI investment may be arriving faster than the supply side can respond.
He also indicated that AI could raise the so-called neutral interest rate in the near term rather than pull it down. In practical terms, that means stronger investment demand may keep underlying economic momentum firmer, even as inflation remains above policymakers’ comfort zone.
Power bills, rate requests, and a growing public backlash
Powell’s remarks land at a moment when the energy consequences of the data center boom are becoming harder to ignore. Utilities have been seeking larger rate increases as they invest in generation, transmission, and grid resilience. Consumers, especially lower-income households, are increasingly exposed to those higher costs through electricity bills.
That has turned data centers into more than a technology story. They are now part of a broader economic conversation about who pays for infrastructure expansion and how quickly public systems can absorb private sector demand. In regions where large clusters of data centers are being developed, concerns about grid strain and household affordability have grown more visible.
The same surge in demand that has excited investors is also creating bottlenecks. Energy supply, permitting, construction capacity, and transmission access are all emerging as limiting factors. That means the inflationary effects are not only about enthusiasm for AI, but also about how difficult and expensive it is to scale the systems behind it.
“We just don’t know” is still the central answer
Powell also said the Fed has been watching unusually strong productivity for several years, something he openly described as surprising. Even so, he stressed that the full economic impact of generative AI remains uncertain. Policymakers can see the investment boom. What they cannot yet measure with confidence is whether productivity gains will outrun the costs being created today.
That uncertainty is central to the Fed’s position. The long-term case for AI remains intact, with the potential to support faster growth and greater efficiency. But Powell’s message was that central bankers cannot make policy based on expected future benefits alone.
For now, the evidence points to an economy where AI investment is boosting demand before it lowers costs. That leaves Jerome Powell data centers inflation as a useful shorthand for a more complicated reality: technological progress may still be coming, but households and businesses are already paying for the buildout.
