The Bank of England simulated a market collapse due to the coordinated behavior of AI agents and other risks from artificial intelligence

The Bank of England is conducting scenario analysis to assess the range of possible macroeconomic and market outcomes from investment in artificial intelligence, its development and implementation, and the associated risks to financial stability. This is stated in the Bank's response to the report of the British Parliamentary Treasury Committee. The letter was signed by Deputy Governor for Financial Stability Sarah Breeden.

The simulations are carried out jointly with international partners. The goal is to understand under what conditions AI agents trading in financial markets may exhibit correlated behavior, or “herding,” and exacerbate a stress scenario. Ways to mitigate such effects are being explored, for example, by adjusting the agents' objective functions so that they take into account public policy goals.

In addition, the Bank is integrating AI scenarios into various forms of cyber resilience and operational resilience testing for the financial sector. Scenario analysis will be part of a broader approach to stress testing, including system-wide exercises. The very idea of ​​AI-specific stress tests was voiced in the April 2025 report of the Bank's financial policy committee and in Breeden's speech in October 2024.

In addition to risks within the financial sector, the Bank is calculating the impact of a sharp fall in the prices of AI-related assets on the entire economy. Back in October 2025, the Financial Policy Committee first identified this problem, and in December it published its analysis. It is noted that while AI infrastructure is financed mainly by the cash flows of large and profitable technology companies and equity investments, debt financing is growing rapidly, including with significant contributions from private markets. The report suggests that if this trend accelerates, the impact of any collapse in AI assets on the UK economy could increase.

Another scenario concerns companies whose business models could be disrupted by AI itself. As noted, fears of this kind against the backdrop of news about rapid innovation in AI have already contributed to sales of shares of IT companies at the beginning of the year. The committee sees the same risks in risky credit markets, where the growing influence of AI on the IT sector has raised concerns about asset quality and the uncertainty of their valuations.

Earlier, the head of the US Ministry of Finance, Scott Bessent, and the Chairman of the Federal Reserve System (FRS), Jerome Powell, held an urgent meeting with the heads of the largest US banks. The reason for this was the release of a new AI model from Anthropic, which the developers presented as a solution with broad capabilities for identifying software vulnerabilities.