Fed Chair Warsh warns superintelligence could bring major disruption and economic costs
FOX 52 / Wikimedia Commons (Public domain)
The central bank chief is treating advanced AI as a macroeconomic force, with a San Francisco forum on the topic scheduled for November
Federal Reserve Chair Kevin Warsh says superintelligence could cause major disruption and carry real economic costs.
Warsh has called the rapid progress of AI a critical turning point in history. He has argued it could reshape economic landscapes both in the US and globally.
He is expected to expand on that view on November 16, 2026, when he leads a public forum in San Francisco. The event centers on “super intelligence,” the Fed’s term for advanced AI, and what it means for productivity, employment and the broader economy.
The forum sits under the Fed’s Productivity and Jobs Task Force. Warsh is slated to deliver keynote remarks. Panels with industry leaders and scholars will cover how businesses are adopting and implementing the technology.
Warsh tapped Marc Andreessen to chair the AI-focused task force.
The inflation problem hiding inside the AI boom
Warsh’s core concern is timing. He has described the current economy as a rocky transition period.
His argument runs like this. Companies are spending heavily on AI infrastructure right now. That spending creates immediate demand for chips, power, construction and labor, which can push prices higher.
The promised payoff, a jump in productivity that would let the economy produce more without stoking inflation, comes later.
AI, tech, and the markets they move—in one daily briefing.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
The scale of the spending is significant. AI infrastructure has been driving around half of recent capital expenditure growth, according to the research findings on Warsh’s agenda.
At Jackson Hole earlier this year, Warsh pointed to another figure that captures the frenzy. Annualized token sales at leading AI labs have topped $100 billion, growing more than 500% year-over-year.
Those tokens are not crypto assets. In AI, a token is a small unit of text that a model reads or writes, and labs sell access to their models by the token. So the figure is essentially a measure of how much the world is paying to use AI systems, and that bill is growing fast.
To get a better read on whether the payoff is arriving, Warsh has asked for real-time data. He wants to know whether a “productivity miracle” can show up quickly enough to offset the inflation and capital spending pressures building now.
A different kind of Fed chair conversation
Warsh took the chair in May 2026, succeeding Jerome Powell. His tenure has quickly become defined by an unusual focus for a central banker.
The superintelligence comment pushes that framing further. Warsh is not only talking about chatbots making office workers more efficient. He is flagging the possibility of disruption large enough to carry economic costs, which is a notably sober note from the person steering US interest rates.
The Fed’s mandate covers both stable prices and maximum employment. AI potentially touches both: it could lift prices in the short term through the investment boom and reshape the labor market over the longer term.
What this means for markets and policy
For investors, the key question is how Warsh’s AI thesis feeds into interest rate decisions. If he believes AI spending is adding to inflation now, that argues for caution on rate cuts.
That is why his request for real-time data matters. The traditional statistics the Fed relies on often arrive with a lag and get revised.
The November forum is the next checkpoint. Watch for whether Warsh leans harder into the inflation risk or the productivity upside, and for what Andreessen’s task force recommends.
For now, Warsh is betting that understanding AI is part of understanding the economy. Given that AI infrastructure accounts for around half of recent capex growth, that is a hard bet to argue with.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.

