Back to homepage

Fed stability report flags data center leverage as Barr examines AI risks

November 20, 2026

On November 13, the Fed released its semiannual Financial Stability Report, which is an assessment of vulnerabilities in the U.S. financial system across four categories using data available as of October 28. The report described the banking sector as “sound and resilient” overall, though it noted sizable fair value losses on fixed-rate assets. Across the four categories, the report found that: (i) asset valuation pressures remained elevated, driven in part by what it characterized as market enthusiasm for “super intelligence”; (ii) vulnerabilities from business and household borrowing were moderate; (iii) vulnerabilities from financial-sector leverage remained notable, with hedge fund leverage at its highest level since the data series began in 2013; and (iv) funding risks were moderate. The report also examined financing for the data center buildout, which it termed “advanced computing” debt. According to the report, issuance of this debt surpassed $680 billion by October, about five times the 2023 total. The report found that other obligations of large technology firms, such as not-yet-commenced leases and guarantees, have grown even faster than their debt. It described the links among these firms, public credit markets, private lenders, and special purpose vehicles as a vulnerability requiring close monitoring.

In a November 16 speech at an international conference on AI in finance, Fed Governor Michael Barr reviewed the report’s findings and discussed other potential financial stability implications of AI. He cited the report’s estimate that the potential maximum nominal value of the firms’ other obligations exceeds $3 trillion. Barr stated that the still-limited debt backing the buildout made him less concerned about direct losses, but he cautioned that, with hedge fund leverage and broader risk appetite “apparently quite high,” a sharp reversal could lead to rapid deleveraging and, potentially, “self-reinforcing fire sales.” He also addressed potential risks from AI-driven trading, AI-enabled cyber threats, and reports from major banks of AI agents attempting to act in retail customer accounts without the banks being able to verify authenticity or consent. On fair lending, Barr cautioned that AI can function as a “black box,” “making it difficult to explain adverse decisions or ensure compliance with fair lending laws,” and that alternative data sources “may also inadvertently serve as proxies for protected characteristics like race or national origin.”