Cook says AI is adding to inflation now and may ease it only later
Victoria Shinder2 min readFederal Reserve Governor Lisa Cook set out how she thinks about artificial intelligence and the Fed's dual mandate in a speech at Oakland Tech Week on 28 September. Her conclusion for the short term was direct: AI appears to be adding inflationary pressure and postponing the return of inflation to the 2% target.
Cook drew a line between two kinds of pressure. Prices of AI-related goods such as chips, computers and software have surged, but she described much of that as a shift of demand into one sector, which should resolve as supply adjusts and which monetary policy should not try to fight, since its tools are too blunt for narrow sectors. The broader concern is spillover: data-centre investment draws on construction labour and energy used across the economy, companies have spent only a small fraction of about $2 trillion in announced plans, and AI-driven equity gains appear to be feeding household spending. She pointed to electricity and water costs, each up around 5% over the past year, as possible signs of broadening.
On the other side, she expects productivity gains to bring modest disinflation within the next few years, but not in time to offset the pressure she expects later this year. The relief is limited, she said, because higher productivity also raises demand through expected wages, returns and wealth. On jobs, she sees limited evidence so far of AI reshaping the labour market, with unemployment and layoffs low, but noted weaker demand in coding and simultaneous translation and harder first-job searches for graduates. A rise in unemployment driven by a skills mismatch, she said, would leave the Fed with limited tools.
Cook confirmed she voted with the rest of the FOMC for the 25 basis point increase at the September meeting. She cited total inflation of an estimated 3.8% in the 12 months to August and core inflation of 3.4%, and unemployment of 4.1% in August.
What it means
The speech gives a framework the market can test against data: sector prices that should fade on their own, and broad prices that should not. The markers she named, utility costs and core goods, are the ones to watch for which of the two is winning.