Federal Reserve Officials Say AI-Driven Productivity Could Lift the Neutral Rate

Several senior officials at the Federal Reserve have indicated that productivity improvements associated with artificial intelligence are likely to raise the economy's neutral rate (r*), narrowing the scope for monetary easing. Governor Michael Barr explicitly warned that a durable AI-driven output uplift should not be read as a reason to lower policy rates, underscoring the Fed's technical distinction between supply-side gains and a justification for looser policy.

The mechanism officials describe is straightforward: if AI increases total-factor productivity and potential output persistently, the equilibrium real interest rate that equilibrates saving and investment will rise, lifting the nominal neutral rate consistent with 2% inflation. That would tighten financial conditions at any given nominal rate, prompting markets to revise expectations for long-run Treasury yields, mortgage costs and corporate discount rates. For macro modelers, the implication is to treat technology shocks as more persistent or to raise steady-state assumptions for potential output and r* in staff projections.

This Fed assessment sits in tension with alternative narratives. Some policy figures and a high-profile Fed nominee argue that rapid AI-driven efficiency could be disinflationary, creating room for meaningful rate cuts if confirmed shifts in prices and wages materialize. Any move toward easier policy faces operational and sequencing challenges — from managing reserve abundance and monthly bill purchases to coordinating with large near-term Treasury issuance — that can complicate the transmission of nominal rate changes into money markets.

European central bankers, including leaders in the ECB, have echoed the productivity potential of AI while stressing that gains are neither automatic nor evenly distributed. They emphasize that outcomes depend on complementary public investment in digital infrastructure, broad-based capital reallocation, and large-scale workforce reskilling; absent those, productivity gains may cluster among a few firms and fail to boost widespread wage income. Those distributional and concentration risks — such as heavy capital spending on compute and data centers concentrated among hyperscalers — could blunt the consumer-demand channel and weaken the transmission of productivity into durable disinflation.

For U.S. policymakers, the upshot is twofold: faster productive capacity can improve supply-side fundamentals but may simultaneously raise the natural rate that anchors policy. That duality complicates the timing and scale of rate reductions and elevates the importance of Fed communication to avoid markets mispricing the permanence of AI gains. At the same time, fiscal, regulatory and labor-market policies will shape whether AI-induced productivity translates into broad-based growth or narrower corporate profits.

Ultimately, the macro effect hinges on the pace, breadth and persistence of AI adoption — plus the public- and private-sector responses that enable diffusion across firms and workers. If productivity gains are broad and sustained, r* and long-term yields could be materially higher; if gains are uneven or transitory, the argument for easier policy would regain traction. Officials say they are monitoring these channels closely and expect models, forecasts and market pricing to adjust as new evidence on the nature of AI-driven productivity arrives.