Sarah Breeden, the Bank of England's Deputy Governor for Financial Stability, has issued a stark warning regarding the potential for AI agents to cause market meltdowns if their use becomes prevalent across financial markets. She highlighted that these AI agents, while potentially offering efficiency gains, could lead to unforeseen vulnerabilities due to their ability to execute correlated actions instantly and on a massive scale. This could result in rapid, unpredictable market shifts that human participants might struggle to comprehend or control, posing a significant risk to financial stability.
Breeden specifically pointed out the risk of "herding" behavior, where multiple AI agents, designed with similar objectives or data sets, might trigger a simultaneous sell-off in response to specific market signals. This concern echoes previous warnings from the Bank of England about AI's potential to amplify market volatility and destabilize the financial system. The central bank has been actively researching and conducting stress tests to understand these risks better, particularly in areas like private credit markets and broader economic shocks caused by geopolitical events like the ongoing war in Iran.
The Bank of England is not alone in its concerns. Global regulators and financial institutions are grappling with how to manage the burgeoning influence of AI in finance. Governor Andrew Bailey previously warned about the unprecedented cyber risks posed by advanced AI products, such as Anthropic's Mythos AI, which could uncover severe cybersecurity vulnerabilities. The Treasury Committee has also criticized the government for delays in bringing AI and cloud companies under the Critical Third Parties regime, which aims to regulate critical suppliers to the financial system, emphasizing the potential for widespread disruption from an outage at a major provider.
Breeden's remarks align with a broader sentiment within financial policy circles that the rapid advancement and adoption of AI necessitate proactive regulatory measures. The Bank of England is working with international counterparts to develop strategies that mitigate these risks, including specific stress testing for scenarios where AI-driven herding behavior could amplify market selloffs. The goal is to ensure that the financial system remains resilient as AI technology integrates more deeply into trading and operational functions, preventing a crash triggered by autonomous agents operating at machine speed.
