Fed Models an AI Boom That Could Cost Jobs and Sink Stocks
The Federal Reserve has started modeling how AI could remake the U.S. economy, and Chair Kevin Warsh laid out three paths Monday in San Francisco: one where workers share the gains, one where automation outruns them, and one where AI spending flops and drags stocks down. He stressed they are scenarios, not forecasts.
- 3 sources report it: decrypt.co, finance.yahoo.com, timesofindia.indiatimes.com.
- The source published it on Nov 16, 2026 at 6:36 PM.
Nothing open. Every check passed.
Federal Reserve Chair Kevin Warsh said Monday that central bank staff have begun modeling how artificial intelligence could reshape the U.S. economy, sketching out three possible futures at the Fed's Productivity and Jobs conference in San Francisco. Warsh said the work tests what might happen over the next several years, and he was blunt about its limits: “These are illustrative scenarios, not forecasts.” The Fed, he said, is preparing for different outcomes rather than predicting which one lands.
In the sunniest version, AI lets companies produce more with the same resources, pushing national economic output above its historical growth rate. Workers share in those gains, and unemployment holds close to the level economists tie to a healthy job market. In the second path, the economy still speeds up, but businesses automate jobs faster than displaced workers can find new ones — so unemployment climbs even as productivity gains pile up.
The third scenario is the ugly one. Companies pour money into AI infrastructure, stock prices climb on the promise of huge profits, and those expectations turn out to be too rich. “Equity valuations could fall, leading to an economic slowdown,” Warsh said. In that version, unemployment rises more sharply and the economy eventually recovers without the productivity boost that was supposed to pay for it all.
Warsh repeatedly used the phrase “super intelligence,” or SI, which the source reports is now official Fed terminology after a change in October tied to a Donald Trump administration executive order — a naming decision, not a scientific finding that today's AI systems already outthink humans.
The risks are drawing scrutiny from elsewhere inside the Fed, too. Governor Michael Barr warned Monday that the AI construction boom is increasingly financed through complicated arrangements, including long-term computing leases, energy contracts and financial guarantees. The Fed pegs the potential maximum value of those off-balance-sheet obligations at more than $3 trillion. They aren't debt already owed, but they show how disappointing AI revenue could ripple well past the companies building data centers.
Warsh floated no interest-rate change in Monday's remarks. The Fed's next policy meeting is set for December 8 and 9.
TopicsKevin Warsh · Federal Reserve · Michael Barr · Donald Trump · San Francisco
Written by AI from verified reporting.

