Bank of England chief warns new AI models threaten global financial stability - CNBC

**Article Body**

The Bank of England’s governor has issued a stark caution to the financial sector, identifying the rapid adoption of advanced artificial intelligence models as a fresh source of systemic risk. While AI promises efficiency in fraud detection and trading, the central bank’s concern centers on a dangerous paradox: the very tools designed to manage risk may now be amplifying it.

The core issue lies in **herding behavior**. If major financial institutions rely on the same third-party AI models or similar training data, their trading algorithms could react identically to a market shock. This synchronized response, often called a "flash crash" dynamic, can drain liquidity faster than human traders can intervene. Furthermore, the opaque nature of some models makes it difficult for regulators—or even banks themselves—to audit the logic behind sudden, large-scale asset movements.

For market participants, this signals a shift in regulatory focus. Expect future stress tests to include scenarios where AI models fail or act unpredictably. The watch point is not just the capability of these systems, but their **concentration risk**—how many critical decisions flow through a handful of unregulated providers. The stability of the financial system may soon depend less on human judgment and more on the resilience of the code that drives it.