🔍 Read the full analysis: Understanding AI’s 87% Showing In The Fed’s Risk Survey on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
In a survey released with its November 2026 Financial Stability Report, the New York Fed said 87% of 23 market contacts cited AI as a risk to U.S. financial stability, compared with 50% of 20 contacts in the spring. The responses reflect those contacts’ views, not the Fed’s own risk ranking; the report separately discusses AI-related cyber risks and borrowing tied to advanced computing.
The Federal Reserve’s November Financial Stability Report says 87% of 23 market contacts surveyed in September and October named artificial intelligence as a risk to U.S. financial stability, up from 50% of 20 contacts in the spring. The result makes AI the survey’s most frequently cited risk, but it is a record of respondents’ views—not a conclusion or probability estimate issued by the Federal Reserve.
The New York Fed collected responses from professionals at broker-dealers, banks, investment funds and advisory firms. It asks contacts twice a year which shocks could cause the most damage to the financial system over the following 12 to 18 months. The report says its information reflects conditions through October 28, so the survey predates developments later in November.
AI led the list, followed by a correction in risk assets, cited by 52% of contacts, and higher long-term interest rates, cited by 48%. Cyberattacks and fiscal debt sustainability each received 43%. Compared with the spring survey, AI rose from 50%, risk-asset corrections from 35%, long-term rates from 20%, cyberattacks from 20%, and fiscal debt sustainability from 30%. The panels differed in size, and the percentages describe the contacts surveyed, not the wider financial sector.
Respondents grouped several concerns under AI. They cited the possibility that AI-related shares could fall if large investments fail to produce expected returns, with high valuations, debt and complex financing potentially adding pressure. They also raised cyber risks, including attacks made easier by AI and possible unintended damage from frontier models, as well as risks from complicated AI-linked financing.
Explained in motion
Banks, insurers and funds often rely on the same few models, platforms and service providers. Supervisors warn that one problem there can reach all of them at once.
Illustrative sequence, not a real incident
Shared providerMany institutions run on the same model, platform or service provider.
One changeA failure, a compromise or a model update changes how that provider’s AI works.
Many hit at onceEvery institution that depends on it is affected at the same time.
Services stallPayments, claims or trading slow down. The Fed warns that disruptions can hurt market liquidity and confidence.
No quick switchThe damage grows when there is no ready substitute for the service.
Source: APRA, System Risk Outlook (November 12, 2026) and letter to industry on AI (April 30, 2026); Federal Reserve, Financial Stability Report (November 2026), pp. 47–48. Illustrative sequence built from the mechanisms they describe.
Graphic: Powered by Thorsten Meyer AI
How AI Risks Could Reach Finance
The survey matters because the concerns span both financial markets and operational security. If expected returns from large AI investments disappoint, a reassessment of related companies could affect investors and lenders. The Fed’s report says borrowing tied to advanced computing exceeded $680 billion by October 2026, about 31% of investment-grade nonfinancial corporate issuance, compared with less than 6% in 2023. These figures describe issuance tied to the buildout; they do not establish that the borrowing will lead to losses.
Cyber risk can spread through shared infrastructure. The Fed warns that an attack on a critical third-party provider could affect multiple institutions, particularly if they cannot readily replace its services. It says disruption at a financial institution could impair market liquidity, weaken investor or depositor confidence and prompt forced asset sales. Those are potential consequences described by the report, not events it says have occurred because of AI.
The report also says the largest U.S. bank holding companies have close to $1 trillion in exposure to firms and vehicles behind the advanced-computing buildout. That figure signals the scale of connections supervisors are watching; the report does not characterize all of that exposure as direct AI lending or as a likely loss.
Products Worth Considering
Bose QuietComfort Wireless Noise-Cancelling Over-Ear Headphones, Cypress Green
Jackery Solar Generator 1000 v2 with 100W Solar Panel
Samsung 32-inch Odyssey G5 G51F QHD 180Hz Gaming Monitor
As an affiliate, we earn on qualifying purchases.
From Survey Concern to Supervisory Review
The Fed’s survey and its own analysis are related but distinct. The survey summarizes opinions from a small, unnamed group of market contacts. Separately, the Board’s report describes the banking system as “sound and resilient”, while noting high equity valuations partly linked to enthusiasm for super intelligence. It also discusses advances in large language models and agentic systems that can identify and exploit software vulnerabilities.
The report says accounts of agentic systems being involved in cyber events have increased since its May report, but gives no count or breakdown. Its warning focuses not only on financial institutions but also on market infrastructure and outside service providers. Dependence on a small number of providers could make disruption harder to contain if customers lack substitutes.
A separate supervisory development came from Australia. In a System Risk Outlook published November 12, prudential regulator APRA named frontier models from Anthropic and OpenAI in discussing improved capabilities to find and exploit software flaws. APRA said its 2027 System Stress Test will examine AI-driven economic disruption. The U.S. survey and the Australian outlook are separate assessments by different authorities, not a joint finding.
Limits of the Survey Evidence
The survey includes 23 contacts, and the Fed does not identify them or say they represent all financial firms. A rise from 50% to 87% indicates more respondents in this panel cited AI; it does not show that the probability of a financial crisis rose by the same amount. The report also does not specify how much of the increase reflects new concerns, changing respondents or both.
Several risk measures remain unquantified. The report gives no count for the reported increase in agentic-system involvement in cyber events, and it does not identify particular incidents or attribute them to a specific model. Nor does it provide a forecast of potential losses from the advanced-computing debt or say that a correction is imminent. APRA’s statements about frontier models and its planned stress test are separate from the Fed’s survey findings.
The Next Supervisory Tests
The Fed report does not announce a specific new rule or timetable for AI oversight. Its findings put attention on cyber preparedness, third-party dependencies and financing, but the report does not say what regulatory action, if any, will follow. Future Fed stability reports and surveys can show whether these concerns persist and whether the panel’s risk rankings change.
APRA’s planned 2027 System Stress Test is a stated next milestone for examining AI-driven economic disruption in Australia. The source material does not give the test’s detailed scenarios, participating institutions or publication date. For now, supervisors and financial firms will be watching how AI-related investment, borrowing and dependence on shared technology providers develop.
Key Questions
What does the 87% figure measure?
It is the share of 23 market contacts surveyed by New York Fed staff in September and October who named AI as a risk to U.S. financial stability over the next 12 to 18 months. It is not a measured probability of a crisis.
Does the figure represent the Federal Reserve’s own view?
No. The report says the survey box summarizes respondents’ views and should not be interpreted as representing the views of the Federal Reserve Board or the New York Fed. The report separately gives the Board’s assessment of financial stability.
What AI-related risks did contacts identify?
They cited possible corrections in AI-related stocks if investment returns disappoint, AI-enabled cyberattacks and complex financing connected with AI. The Fed’s report separately warns that attacks on shared service providers could affect multiple institutions.
What is the significance of the $680 billion figure?
The Fed reports that debt issuance tied to advanced computing exceeded $680 billion by October 2026, amounting to about 31% of investment-grade nonfinancial corporate issuance. The figure indicates the scale of financing activity; the report does not say it will result in losses.
What happens next?
APRA says its 2027 System Stress Test will examine AI-driven economic disruption. The Fed report sets out no specific new rule or follow-up date, so further supervisory steps are not yet clear.
Source: ThorstenMeyerAI.com
