Federal Reserve governor Lisa Cook warned that data-centre construction may raise demand for electricity and building labour used across the economy. She distinguished those pressures from price rises confined to AI chips and software, which monetary policy cannot target neatly. Cook expects productivity benefits over the next few years, but said they may arrive too late to offset wider inflation pressure later in 2026. She did not assign all recent utility-price increases to AI.
Federal Reserve governor Lisa Cook warned that artificial-intelligence investment could spread price pressure beyond AI equipment. Data centres need electricity and construction labour, inputs used by many other businesses. Demand for those resources can therefore matter more broadly than higher prices for chips or software alone. Cook framed this as a risk, not a measurement of how much current US inflation is attributable to AI.[1], [2]
Cook distinguished that possible spillover from price increases confined to AI chips, computers and software. A shift of buyers toward one sector can lift its prices even when inflation has not spread through the rest of the economy. She said monetary policy acts on overall demand and cannot neatly suppress one sector’s relative price change. Her distinction matters because the policy response to a narrow price movement differs from the response to broad price pressure.[1]
Cook said companies have spent only a small fraction of roughly $2 trillion in announced AI investment plans. She expects productivity gains to ease some inflation pressure over the next few years, but doubts they will arrive soon enough to offset broader pressure later in 2026. She cited rises of about 5% in electricity and water costs over the past year without assigning all of either increase to AI. The timing and size of any productivity benefit remain uncertain.[1]
