Fed Chair Warsh Outlines AI 'Super Intelligence' Scenarios, Cites $3 Trillion in Off-Balance-Sheet Risks
Summary: Federal Reserve Chair Kevin Warsh described three illustrative economic paths for artificial intelligence, ranging from shared productivity gains to an investment bust. Governor Michael Barr warned that complex financing for AI infrastructure has created over $3 trillion in potential off-balance-sheet obligations.
Federal Reserve Chair Kevin Warsh announced on Monday that central bank staff have begun modeling how artificial intelligence could reshape the U.S. economy, including scenarios where the technology raises unemployment even as overall economic growth accelerates. Speaking at the Fed's Productivity and Jobs conference in San Francisco, Warsh described the initiative as a preparatory measure rather than a prediction of future outcomes. Warsh said Fed economists are modeling how AI “super intelligence” might affect economic growth and employment, but stressed that the scenarios are not forecasts.
Three Illustrative Scenarios
Warsh detailed three distinct paths that the adoption of advanced AI technologies could take over the coming years. In the most optimistic scenario, AI enables companies to produce more with the same resources, raising the country's economic output above its historical growth rate. In this outcome, workers share in the gains, and unemployment remains close to the level economists associate with a healthy job market. At one end, SI leads to an increase in productivity and economic gains that are shared broadly across labor and the owners of capital.
The second scenario presents a more complex dynamic where AI still drives faster economic growth, but businesses automate jobs more quickly than displaced workers can find new roles. This path results in rising unemployment despite broader productivity gains. Warsh noted that widespread automation, combined with labor market frictions, could push the jobless rate above levels consistent with full employment. One outcome brings stronger growth and broad gains for workers; another delivers growth alongside higher unemployment as automation replaces jobs.
The third scenario envisions an AI investment bust. In this case, companies spend heavily on infrastructure and stock prices rise on expectations of huge profits, but those expectations eventually prove too optimistic. Warsh warned that equity valuations could fall, leading to an economic slowdown and a sharper rise in unemployment. The economy would eventually recover, but without the promised productivity boost. The third envisions disappointing AI returns, falling stock valuations and an economic slowdown that could also pressure riskier assets such as crypto.
Official Terminology and Regulatory Context
During his remarks, Warsh referred to the technology as “super intelligence,” or SI. This terminology is now official Federal Reserve language following an update in October. The change was made in response to a Donald Trump administration executive order issued in September 2026, rather than reflecting a scientific determination that current AI systems have surpassed human intelligence. On October 30, 2026, the Board updated references from artificial intelligence (AI) to super intelligence (SI) where appropriate based on the September 29, 2026, Executive Order 14434 on Inaugurating the Era of Super Intelligence.
Warsh emphasized that the Fed’s role is neither advocacy nor opposition, but rather recognizing the significant impact widespread adoption will have on the economy. He stated that the central bank is using simplified economic models to test what might happen over the next several years. “These are illustrative scenarios, not forecasts,” Warsh said, explaining that staff are using simplified economic models to test what might happen over the next several years.
Financial Risks and Off-Balance-Sheet Obligations
Beyond the macroeconomic scenarios, Federal Reserve Governor Michael Barr highlighted specific financial risks associated with the AI construction boom. Barr warned that this expansion is increasingly financed through complex arrangements, including long-term computing leases, energy contracts, and financial guarantees. These structures create potential vulnerabilities that extend beyond the technology companies building data centers. Governor Michael Barr warned Monday that the AI construction boom is increasingly financed through complex arrangements, including long-term computing leases, energy contracts and financial guarantees.
The Federal Reserve estimates that the potential maximum value of these off-balance-sheet obligations now exceeds $3 trillion. While these commitments are not the same as debt already owed, they indicate why disappointing AI revenues could create broader economic problems. The Fed estimates that the potential maximum value of these off-balance-sheet obligations now exceeds $3 trillion.
Warsh did not propose any interest-rate changes during his Monday remarks. The Federal Reserve’s next policy meeting is scheduled for December 8 and 9, where officials will continue to assess economic data and emerging risks. Warsh did not propose an interest-rate change in Monday's remarks. The Fed's next policy meeting is scheduled for December 8 and 9.