Artificial intelligence is throwing sand in the gears of monetary policy. In a study from the Bank for International Settlements, Leonardo Gambacorta and Salvatore Polizzi argue that changes tied to AI - spanning productivity, how well economies withstand climate stresses, and limits on energy - are reshaping potential output, thereby clouding signals from capacity utilization and inflation.
One wrinkle they highlight: AI could lift energy costs before efficiency gains show up, making it harder for policymakers to judge the balance between demand and supply.
The real cost of AI infrastructure is harder to measure than it looks. Market Briefs covers that question free every morning.
The study says the net climate impact of AI depends on two things. First, whether efficiency improvements outrun the extra electricity demand from data centers. Second, how that power is generated.
As the authors put it, "The coming decade will be critical in determining whether AI becomes a force that accelerates the transition to a more sustainable economy or one that intensifies existing environmental pressures."
The paper also flags financial stability concerns. It warns that pouring significant capital into data centers and electricity grids could spark abrupt corrections if expectations prove overly optimistic.
Those cautions echo the BIS's recurring warnings that a bust following the AI surge could inflict broad damage. In a September speech, Pablo Hernandez de Cos, the BIS chief, said AI company valuations are built on "ambitious expectations" and have sparked an "arms race" among them.
Energy and water use are becoming part of the investment case. Get the free Market Briefs daily newsletter and follow the research.
