Business

AI Becomes the Top Risk in the Fed's Financial Stability Survey as Data Center Debt Passes $680 Billion

Eighty-seven percent of the market contacts the Federal Reserve polled named artificial intelligence as a salient threat to the financial system, up from 50% in the spring. The same report puts new numbers on the borrowing, leases and guarantees behind the data center boom.

DailyMech Staff

Published

Last updated / Fact-checked on

AI Becomes the Top Risk in the Fed's Financial Stability Survey as Data Center Debt Passes $680 Billion
PIX1861 / Pixabay — Pixabay Content License

Close-up of red and green candlesticks on a dark stock-market chart screen

Illustrative photo of a candlestick price chart. It does not show any data from the Federal Reserve report.

Artificial intelligence is now the most-cited worry among the Wall Street contacts the Federal Reserve asks about threats to the financial system, and it is not close.

In its twice-yearly Financial Stability Report, released on Friday, Nov. 13, the Fed said 87% of the contacts surveyed in September and October named AI as a salient risk to U.S. financial stability. In the spring survey, published in May, the figure was 50%. The next most-cited risk, a correction in risk assets, was named by 52%.

The report also goes beyond the survey. In a box titled "Leverage in the Data Center Buildout," Fed staff lay out how quickly borrowing for AI infrastructure has grown, and how much of the industry's exposure sits off balance sheets in leases, guarantees and contracts.

At a glance

Item Detail Source
Released Nov. 13, 2026 Federal Reserve Board
AI named as a salient risk 87% of contacts, up from 50% in the spring Fed report; Reuters; CNBC
Other top risks Risk asset correction 52%; higher long-term rates 48%; cyberattacks 43%; fiscal debt sustainability 43% Fed report
Survey 23 contacts, September and October Fed report
AI-related debt issuance More than $680 billion by October 2026, versus $130 billion in all of 2023 Fed report
Share of investment-grade issuance About 31% in 2026, up from under 6% in 2023 Fed report
Big-bank exposure Close to $1 trillion to tech firms, data center developers, related vehicles and utilities Fed report

What the survey found

The survey is run by staff at the Federal Reserve Bank of New York, who collect views from broker-dealers, banks, investment funds and advisory firms. This round drew responses from 23 contacts. The Fed stresses that the box summarizes respondents' views and does not represent the views of the Board or the New York Fed.

Respondents described three separate AI worries. In the report's words, they were "(1) elevated equity valuations and the possibility that lower-than-expected earnings could result in lower valuations and a market correction; (2) cybersecurity risk, stemming from either a malicious attack by sophisticated state actors or damage caused by frontier AI models themselves; and (3) risks associated with complex forms of AI-related financing."

The other top risks lean on the same theme. Contacts who cited a risk-asset correction pointed specifically to AI-related stocks, "driven by uncertain returns on large capital expenditures," and said circular financing and heavy debt could amplify losses that spill over into business investment and household spending. Those who cited cyberattacks said advances in AI are lowering the barrier to sophisticated attacks, and some flagged unintended damage from frontier models. Mentions of AI, higher long-term rates and cyberattacks rose the most compared with the spring survey.

The timing matters. Reuters noted that the survey closed before Nvidia chief executive Jensen Huang resigned on Thursday, Nov. 12, and before this month's midterm elections, in which public concern about AI was a central issue. CNBC reported that the report covers the period up to Oct. 28.

The debt behind the buildout

The report's data center box is the more detailed part. Fed staff say debt issuance tied to advanced computing infrastructure rose to more than $680 billion by October 2026, about five times the $130 billion issued in all of 2023. Investment-grade technology companies issued most of it. As a share of nonfinancial investment-grade corporate debt issuance, AI-related borrowing climbed from under 6% in 2023 to about 31% this year. Outstanding public debt related to advanced computing is estimated at more than $1.1 trillion, and the Fed says private credit deals make the true total harder to measure.

The mix is also shifting, the report says, from mostly corporate bonds toward a blend of public and private debt, structured credit and project finance.

Fed staff compare the cycle with the broadband telecom boom that began in 1995. Measured against GDP, five hyperscalers' cumulative capital spending since 2020 has tracked the telecom industry's closely, but the report says AI investment has been accelerating and "may surpass that of the 1990s telecom cycle relatively soon." Hyperscalers' strong cash flows had limited their need to borrow, "at least until this year." The report defines the five as Amazon, Alphabet, Meta Platforms, Microsoft and Oracle.

Off the balance sheet

The faster growth is in obligations that are not counted as debt. For a group of seven companies, Amazon, Alphabet, Broadcom, Meta Platforms, Microsoft, Nvidia and Oracle, the Fed tallies leases that have been signed but not yet started with undiscounted values above $1 trillion, and other commitments such as take-or-pay contracts and residual value guarantees above $2 trillion. The report cautions that these undiscounted figures are not directly comparable with on-balance-sheet debt. Reuters summarized the combined picture as debt, guarantees and leases that could exceed $3 trillion.

The report explains how these tools work. A hyperscaler signs a long-term lease with a data center developer, and the developer uses that lease to borrow for construction. GPU makers often guarantee the resale value of the chips used as collateral when AI companies buy them. Equity stakes from hyperscalers and chip designers on top create what the Fed calls a "circular financing" dynamic.

That cuts both ways, the Fed says. The largest firms may absorb losses in a downturn, shielding the wider system. But their real exposure "may be substantially larger than suggested by their on-balance-sheet liabilities." Riskier subordinated slices of these deals sit with high-yield bond funds, life insurers, business development companies and other private debt vehicles. The largest U.S. bank holding companies have close to $1 trillion of exposure to large tech firms, data center developers, related special purpose vehicles and private credit funds, and utilities. Those exposures are covered in the annual stress test.

The report's overview sums it up: "The complex interlinkages among firms engaged in the data center buildout, public credit markets, private lenders, and special purpose financing vehicles create vulnerabilities that necessitate close monitoring."

The rest of the system

Outside AI, the Fed's overall picture is steadier. It describes the banking sector as "sound and resilient overall," with capital near historically high levels. Household debt relative to GDP continued to fall. The sore spot is leverage at hedge funds, which reached the highest level since the data began in 2013 and is concentrated in a small number of large funds.

Why it matters

The Financial Stability Report is not a forecast, and a survey of 23 market contacts is a measure of sentiment, not proof of danger. But the report puts the central bank's own numbers on how the AI boom is being financed, including the large share of commitments that conventional debt measures miss.

For investors, the message is that the central bank is watching the same concentration that has made AI stocks dominate markets. For AI companies and their lenders, the off-balance-sheet tally shows that Fed staff are looking at leases and chip guarantees, not just bonds, when they size up the risk. And for policymakers, the warning lands as data center costs and electricity use have become political issues; CNBC cited exit polling in which 56% of midterm voters called data centers of crucial or major importance to their choices.

How we reported this

DailyMech read the Federal Reserve's November 2026 Financial Stability Report directly, including the overview, the data center leverage box and the survey box, and read Reuters' and CNBC's reports in full. All figures and quotations attributed to the Fed come from the report. The $3 trillion framing is Reuters', and the Oct. 28 cutoff and exit-poll figure are CNBC's. DailyMech contacted no one for this article.

Published November 14, 2026. Facts checked against the cited sources by DailyMech's AI writer.

Sources

DailyMech used AI tools to research and draft this article. Facts were checked against the cited sources by DailyMech's AI writer, not by a human editor.

In Business

From other sections

Sections