• Fed Chair Kevin Warsh warned that disappointing AI investment returns could trigger an equity selloff, economic slowdown, and higher unemployment, while stressing that the AI economic scenarios are illustrative, not forecasts.
  • Warsh said Fed staff have begun modeling economic scenarios for superintelligence, as the central bank assesses how emerging technologies could reshape the economy and monetary policy, with outcomes ranging from broad productivity gains to high unemployment and economic slowdowns.
  • Speaking as the Fed’s Productivity and Jobs Task Force convened in San Francisco, Warsh stressed that humans must retain decision-making authority and warned that AI’s risks could be highly disruptive and economically costly.

A Sobering Assessment

Fed Chair Kevin Warsh warned that emerging technologies, including superintelligence, could fundamentally reshape the economy and monetary policy, offering a stark assessment of the potential outcomes that range from broad productivity gains to high unemployment and economic slowdowns. In remarks tied to the Fed’s Productivity and Jobs Task Force Conference in San Francisco, Warsh said the technology’s risks could be highly disruptive and economically costly, while stressing that humans must retain decision-making authority. He added that the Fed must assess the scale, timing, and economic implications of technological change.

Warsh said Fed staff have begun modeling economic scenarios for superintelligence, while cautioning that the AI economic scenarios are illustrative, not forecasts. He also warned that disappointment in AI investment could trigger an equity selloff, an economic slowdown, and higher unemployment.

Warsh’s comments come as policymakers gather at the San Francisco Fed to examine AI adoption, productivity, employment, and business investment. The conference, which was scheduled to begin Monday morning, brings together technology executives, economists, and employers—including representatives from Stripe, Affirm (NASDAQ:AFRM), and Palo Alto Networks (NASDAQ:PANW), as well as robotics, packaging, and healthcare firms—to explore how technological change is affecting the economy and labor market.

A Shift in the Narrative

The Fed chair’s warning marks a notable evolution in his thinking. Since taking the helm in May, Warsh has moved from a relatively optimistic view that AI would lower inflation to a more complicated assessment: building the technology can raise costs now, even if using it eventually makes production cheaper. The shift reflects a broader debate within the central bank, where officials have grown more vocal about inflation and employment risks.

Richmond Fed President Tom Barkin has warned that the historically large investment wave could become a vulnerability if it fails to pay off and companies scale back. Governor Lisa Cook has cautioned about at least a temporary increase in unemployment, potentially including a higher underlying level of joblessness, while Governor Michael Barr warned of “serious short-term disruptions” depending on deployment speed and affordability. Warsh himself has asked whether the initial economic surplus will accrue mainly to AI laboratories, chipmakers, energy producers, and cloud providers—or ultimately spread to other businesses, workers, and consumers.

The Investment Boom and Its Risks

AI investment is already supporting U.S. growth, stock-market gains, and business and consumer spending, but the scale of the buildout carries risks. Warsh has rejected the idea that exceptionally high expected AI returns make financing costs unimportant, saying rates affect investment throughout the economy. The Fed faces three distinct possibilities: productivity gains arriving broadly and quickly, investment running ahead of productivity and keeping inflation persistent, or investment disappointing and labor displacement accelerating.

The political backdrop adds another layer of complexity. Recent midterm elections brought a backlash against rapid data-center construction, with local disputes over land use, displacement, and whether communities share in the benefits of the investment. Oakland Mayor Barbara Lee has supported opposition to local data-center construction and warned against discrimination and displacement.

Institutional Framework

The conference is part of the Fed’s broader effort to develop evidence for the Federal Open Market Committee. In June, the central bank created five independent task forces, including the productivity-and-jobs group chaired by venture capitalist Marc Andreessen. Their work informs policy but does not itself set interest rates. The November conference is an evidence-gathering exercise, not an announced AI regulation or interest-rate decision.

At Jackson Hole in August, Warsh described AI as potentially a new factor of production and posed questions that remain unresolved: Will productivity rise significantly and persistently? When? And where will the returns accrue? The San Francisco conference is a practical continuation of that inquiry, with the breadth of evidence and competing perspectives a key focus. Barkin has explicitly called for hearing strong arguments on both sides.

The most useful takeaway from Warsh’s remarks is that technological success, investor returns, and broad economic welfare are separate questions. The Fed’s inquiry adds crucial questions of timing, employment disruption, and near-term inflation to a debate that will shape monetary policy for years to come.