Federal Reserve governor Lisa Cook identified the artificial-intelligence infrastructure buildout as a major inflation risk for 2027 at an event at the Federal Reserve Bank of New York on 1 October 2026. Speaking to New York Fed president John Williams, she warned that the pressures it creates might persist, according to a Reuters report.
Key points
- Cook sees the AI buildout as a principal inflation concern for 2027.
- She expects AI to improve productivity over the longer term, but worries about when any disinflationary gains will arrive.
- Cook also raised the possibility that persistent supply shocks could call for a different monetary-policy response, depending on the sectors affected.
Cook’s concern is the AI buildout
Cook’s concern centres on the period in which AI infrastructure is being built. “The AI build out is potentially creating inflationary pressures that may not resolve very quickly,” she told Williams. She called it “one of the main things that concerns me right now for 2027.” Her wording leaves the inflationary effect as a possibility, while making clear the weight she places on that risk in her outlook.
Cook also expects AI to raise productivity over the longer term. That prospect does not settle her immediate concern. “I worry about when the productivity gains that would produce disinflation will come, and where the supply bottlenecks are going to be next,” she said. The question for her is therefore one of timing as well as direction: whether productivity can ease prices soon enough to matter while the buildout is putting pressure on them.
That distinction matters for monetary policy. Cook is treating the possible productivity benefit and the possible inflationary cost as developments that need not arrive together. She expressed confidence in the former over the longer term, but described the latter as a concern for 2027. Neither view requires her to discount AI’s eventual economic contribution in order to worry about prices during its expansion.
The Fed’s 2% goal frames the warning
The warning comes against an established inflation problem. Inflation on the measure targeted by the Fed was 3.4% in August and had been above its 2% goal for more than five and a half years, Reuters reported. Cook’s remarks concern another potential source of pressure while the central bank is already trying to return inflation to that goal.
Cook joined a unanimous Fed vote in September to raise the policy rate by a quarter of a point. The increase was intended to support a “timelier” return to the 2% goal. Her remarks at the New York Fed did not present AI productivity as a near-term answer to the inflation the bank is addressing. Instead, she placed the arrival of any disinflationary gains among the questions facing policymakers.
The distinction between an eventual productivity improvement and current price pressure also limits what can be taken from a favourable long-term view of AI. Cook said she believes the technology will boost productivity. She expressed concern, in the same discussion, about where supply constraints might appear next. Both considerations enter her account of the inflation outlook.
Supply shocks complicate Cook’s policy view
AI was not Cook’s only concern. She said supply shocks had become more frequent and had produced effects that were surprisingly persistent, making them more important for policy. Geopolitical developments, including the conflict in the Middle East, could also restrict supply chains. Those risks sit alongside the AI buildout in her account of the pressures the Fed may have to confront.
Cook described an older approach under which policymakers would generally look through supply shocks. Higher interest rates could not change an oil price or end a war, she said, but could slow employment and output. That trade-off explains why a central bank might hesitate to tighten policy against a disruption to supply rather than a rise in demand.
She now considers it possible that the best response could differ according to which sectors a supply shock affects. That is a conditional position, rather than a prescription for the Fed to react in the same way to every disruption. For an AI buildout that Cook fears may generate lasting price pressure, the question of where bottlenecks arise becomes part of the policy judgement.
