The Federal Reserve’s Eccles Building in Washington, DC, pictured in 2011 (file photo). Image: Federal Reserve / Wikimedia Commons, Public domain, cropped
.jpg){kind=link}
Artificial intelligence is now the risk Wall Street worries about most, according to the Federal Reserve. Its November 2026 Financial Stability Report, published on Friday, says AI was the top-cited risk to US financial stability in a New York Fed survey of 23 market contacts, up from third place in the spring.
From third place to first in six months #
Staff at the New York Fed asked professionals at broker-dealers, banks, investment funds and advisory firms in September and October which shocks over the next 12 to 18 months worried them most. AI came top, ahead of a correction in risk assets, higher long-term interest rates, cyberattacks, US government debt levels and higher energy prices.
In the spring survey of 20 contacts, geopolitical risks led, followed by an oil shock, with AI in third. Mentions of AI, higher long-term rates and cyberattacks rose the most since then. The Fed's own section on near-term risks says "most of the respondents" raised risks tied to "new computing technologies". The report stresses that the survey sums up the respondents' views, not those of the Fed's board or the New York Fed, and the text gives no percentages.
Three ways AI could hit the system #
Respondents named three separate worries. The first is valuations: that "lower-than-expected earnings could result in lower valuations and a market correction". Many said "stretched valuations, circular financing arrangements, and heavy reliance on debt to fund the AI buildout could amplify losses", which could spill into business investment and household spending.
The second is cyber risk, "stemming from either a malicious attack by sophisticated state actors or damage caused by frontier AI models themselves". Many said AI is "lowering the barrier to sophisticated attacks", and the Fed adds that reports of agentic AI systems being involved in cyber events have become more frequent since May. That is a new note: in the spring, AI worries were about valuations, debt-funded spending and job losses. The Bank of England raised rogue AI as a cyber risk in September, and the G20's Financial Stability Board has told banks to give every AI agent an ID and a kill switch.
The third is complex AI-related financing, which the Fed examines in a box of its own.
More than $680 billion of new AI debt #
The Fed says debt issued to pay for advanced computing passed $680 billion by October 2026, about five times the $130 billion issued in all of 2023. It now makes up about 31% of new investment-grade debt issued by non-financial companies, up from under 6% in 2023, and public debt outstanding is put at more than $1.1 trillion, with more hidden in private credit.
It also points to obligations that sit off balance sheets: long data centre leases that developers use to borrow, guarantees on the resale value of GPUs used as collateral, and equity stakes on top, which it says create a "circular financing" dynamic. The biggest US banks have close to $1 trillion of exposure to tech firms, data centre developers, financing vehicles and utilities. The Fed says spending by Alphabet, Amazon, Meta, Microsoft and Oracle may soon outpace the 1990s telecom boom. Elsewhere it puts high share prices partly down to "market enthusiasm for super intelligence (SI)", the term Trump ordered federal agencies to use.
A warning in a rough month for AI stocks #
The report lands after the Nasdaq fell into a correction on November 5, a sell-off tracked in our AI bubble guide, along with the IMF's October warning about an AI market correction. Nvidia is run by an interim chief after Jensen Huang resigned on Thursday, and Anthropic is due to price an IPO that values it at up to $1 trillion.
Why it matters #
The Fed is not saying a crash is coming, and the banks it supervises are well capitalised. But the people who trade and lend for a living now rank AI above war, oil and inflation as the thing most likely to break something, and the Fed's own numbers show how much debt now rides on the AI build-out paying off.
Sources: Federal Reserve Board, Financial Stability Report (November 2026); Federal Reserve Board, Financial Stability Report (May 2026), near-term risks.