Stay informed with free updates
Simply sign up to the Central banks myFT Digest — delivered directly to your inbox.
The AI boom is blurring the economic signals central banks rely on to set monetary policy, raising the risk they will make damaging mistakes, economists at the Bank for International Settlements have warned.
Staff at the Basel-based organisation, which advises the world’s central banks, said AI’s effect on investment, trade and asset prices was powerful enough “to shape the global outlook in real time”, supporting growth in the face of trade tensions and geopolitical shocks.
These effects were now “large and observable” and could add to price pressures, they said, noting that spending on data centres and IT manufacturing facilities in the US had risen to 0.8 per cent of GDP, while the wealth effect of equity price gains was boosting consumers.
AI could also have disinflationary effects, however, if it improved productivity — or if worries about AI-related job losses curbed spending and reduced workers’ bargaining power.
Assessing the extent and timing of these effects was a challenge for central banks, the BIS economists warned, because strong GDP growth could simply reflect the temporary investment boom and wealth effects, while more lasting productivity gains were “uncertain and hard to measure”.
“The relative strength and timing of these forces remain uncertain,” they said in an analysis published in BIS’s latest monthly bulletin on Tuesday, adding that “near-term inflationary effects may already be emerging” while “disinflationary effects are likely to emerge more gradually”.
Greater uncertainty increases the danger of monetary policy “miscalibration”, the BIS staff said, arguing that if central banks overestimated productivity gains or underestimated the rise in underlying demand, they risked leaving interest rates too low to contain inflation.
Their intervention came as the US Federal Reserve began a two-day meeting to set policy against a backdrop of growing concern about the extent to which the AI boom is stoking price pressures in the American economy.
Fed chair Kevin Warsh has argued that the US is on the cusp of an AI-induced productivity renaissance that will give the Fed room to cut borrowing costs without fuelling inflation.
But some US rate-setters warn that in the near term, investment in data centres and demand for AI-associated products are leading to higher prices. With inflation running at more than double the Fed’s target on its preferred PCE measure, Warsh is under growing pressure to show the central bank is serious about restoring price stability.
Other central banks are also grappling with the uncertainties of AI.
Philip Lane, chief economist at the European Central Bank, said earlier this month that its effects would depend on whether the technology displaced workers or helped them produce more and on whether energy supply could keep up with rising demand for power. He also cited other uncertainties, including whether AI activity would remain concentrated in the US and China and the speed of adoption.
Assessing AI’s overall impact on inflation would be “a major challenge” for central banks in the years ahead, Lane said.
