Artificial intelligence’s potential to boost productivity may not translate into lower inflation, according to research published by the Bank of England’s Bank Underground blog. The paper, co-authored by International Monetary Fund chief economist Silvana Tenreyro, argues that early-stage AI investments could exacerbate inflationary pressures before productivity gains are realized.
The study highlights a disconnect between investment in AI infrastructure and actual output improvements. Tenreyro, a former Bank of England policymaker, notes that business and household spending often outpace realized productivity gains—a dynamic she suggests is already unfolding with AI-related capital expenditures. Such front-loaded demand can strain supply chains, particularly in sectors like electronics, where prices for components such as memory chips and graphics processors have surged over the past year.
The inflationary impact of productivity gains, the research finds, depends heavily on where they occur. Productivity improvements in services are more likely to reduce domestic inflation, while gains in export-driven sectors may raise wages and increase demand for constrained services, ultimately pushing prices higher. The findings underscore the challenges central banks face in calibrating monetary policy amid evolving technological disruptions.
The paper was published on August 20, 2026, as part of the Bank of England’s Bank Underground series, which features staff research not representing official central bank policy.
