(Washington, agencies) Risks in the artificial intelligence (AI) sector have emerged as the biggest concern for financial stability in the US Federal Reserve's semi-annual Financial Stability Report.
The survey released on Friday (13 November) found that the potential risks posed by AI technology itself and the rapid growth of investment in the sector were respondents' top concerns.
Some 87% of respondents viewed AI as a significant risk to the financial system. Their concerns included cyberattacks, risks associated with complex AI-related financing, and the possibility that slowing earnings growth in the sector could trigger a broad market correction. In the previous survey, released in May, only 50% of respondents identified AI as a risk factor.
A broad correction in risk assets was respondents' second-biggest concern, cited by 52%. Cyberattacks were cited by 43%, higher long-term interest rates by 48%, and fiscal debt sustainability by 43%.
The latest survey was conducted in September and October, meaning respondents had yet to factor in the recent controversy surrounding Nvidia and Jensen Huang.
The report examined the scale of debt issuance linked to data centres and other advanced computing infrastructure. As of October 2026, advanced computing-related debt issuance had surged to US$680 billion (about S$875.3 billion), five times the total issued in the whole of 2023.
Such debt now accounts for nearly a third of all non-financial investment-grade corporate debt issuance, up from less than 6% three years ago.
"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.
The report also noted that investment in AI had been an important driver of US economic growth this year. An unexpected decline in corporate earnings could lead to a market downturn, tighter credit and reduced investment, weighing on the economy.
At a meeting on Thursday (12 November), the AI task force under Donald J. Trump's administration recommended easing cybersecurity liability policies to encourage AI companies to voluntarily report significant incidents. This marked a sharp departure from its earlier tough public stance.
Bloomberg, citing unnamed officials, reported that members of the AI task force believed previous standards for regulatory action over cybersecurity incidents might be outdated and had discouraged the sharing of important information. The task force therefore urged the relevant agencies to limit regulatory liability for the next 12 months or longer.
The latest stance signals that the Trump administration favours offering AI companies the “carrot” of liability exemptions rather than wielding the “stick”, a marked departure from its earlier statements.
The administration had previously indicated that it intended to use existing laws to punish companies that produced unsafe products. As recently as last month, the AI task force warned companies that they must report and address safety incidents or face unspecified consequences.
