Key facts
- Eighty-seven percent of Fed survey contacts cited AI as a salient financial system risk, up from 50% in May.
- The Fed said advanced computing-related debt issuance reached $680 billion by October 2026, five times all of 2023.
- Such borrowing was nearly a third of nonfinancial investment-grade corporate debt issuance, versus under 6% three years earlier.
- The Fed said debt, guarantees and not-yet-commenced leases on AI infrastructure could exceed $3 trillion.
- Hedge fund leverage, especially at the largest firms, remained at the highest levels since Fed data collection began in 2013.
- The survey was conducted in September and October, before Nvidia CEO Jensen Huang stepped down on Thursday.
Artificial intelligence risks became the most-cited concern for financial stability in the Federal Reserve's semi-annual Financial Stability Report, released on November 13, 2026. Eighty-seven percent of contacts surveyed cited AI as a salient risk to the financial system, up from 50% in the May version of the survey, Reuters reported.
Respondents pointed to worries about cyberattacks, risks tied to complex forms of AI-related financing, and the possibility of an AI earnings slowdown prompting a broad market correction. A broad correction in risk assets was the second most-cited concern at 52%, cyberattacks specifically were flagged by 43%, higher long-term interest rates by 48% and fiscal debt sustainability by 43%.
The survey was conducted in September and October, so respondents weighed in before Nvidia CEO Jensen Huang stepped down on Thursday after revelations of extramarital affairs, including one involving a Chinese businesswoman whom the New York Times has reported is under suspicion as a Chinese spy. It also preceded the results of recent midterm elections in which rising public concern about AI was a central issue.
The report itself examined the rapid increase in debt issuance tied to data centers and other advanced computing infrastructure. The Fed found advanced computing-related debt issuance had reached $680 billion by October 2026, already a fivefold increase on debt issued in all of 2023. Such borrowing accounted for nearly a third of all nonfinancial investment-grade corporate debt issuance, after amounting to less than 6% three years earlier.
The Fed compared the pace of debt issuance and investment in AI infrastructure to the broadband telecom investment cycle of the late 1990s, saying the current cycle may soon surpass that era. It focused on financing arrangements such as long-term contracts, leases and guarantees that let hyperscalers and other large firms support projects financed by smaller third parties, with graphics card designers and hyperscalers often contributing equity as well. Fed data showed debt, guarantees and not-yet-commenced leases on this infrastructure could exceed $3 trillion.
The Fed said such circular financing dynamics could let the largest firms absorb losses in a downturn, limiting the impact on the broader financial system, but could also mean their exposure to a downturn is substantially larger than their on-balance-sheet liabilities indicate. "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 report stated.
Elsewhere, the Fed found asset valuations across a range of classes remained high, due in part to investor enthusiasm about AI. It said AI investment has been a significant driver of US economic growth in 2026, and an unexpected downturn in corporate earnings could prompt a market decline, tighter credit and reduced investment. Hedge fund leverage, particularly among the largest firms, remained at the highest levels recorded since the Fed began collecting such data in 2013.