Bank of England Governor Andrew Bailey has issued a warning regarding the rapid expansion of artificial intelligence, stating that a sharp correction in AI-related investments could trigger significant shocks across global financial markets. The warning, formalized in the Financial Policy Committee’s (FPC) latest quarterly record on September 30, 2026, signals a shift in the central bank’s assessment of AI from a potential productivity driver to a systemic financial stability risk.
According to data released by the Bank of England and Morgan Stanley, global AI-related debt issuance reached approximately $450 billion in 2026. This volume represents a doubling of the $225 billion in debt recorded in 2025, reflecting the accelerating pace at which companies are borrowing to fund infrastructure and development.
The Risk of Recursive Learning
A primary concern cited by Governor Bailey is the emergence of “recursive learning” in frontier AI models. This technical phenomenon involves autonomous systems improving themselves in a closed loop, which could eventually allow models to govern their own operations with minimal human oversight. Bailey noted that such autonomy could reduce the ability of regulators or market participants to intervene effectively during periods of financial stress.
The FPC’s analysis suggests that the interconnected nature of these models poses a risk to market transparency. If AI systems begin to manage financial portfolios or execute trades based on closed-loop learning, the logic behind market moves could become opaque, complicating the response to sudden volatility.
Leverage and Market Vulnerability
The Bank of England further highlighted that high valuations in the AI sector are being compounded by increased leverage. A growing number of investors are utilizing borrowed funds to speculate on AI growth, a trend that typically amplifies the impact of any market downturn. The FPC maintained the Countercyclical Capital Buffer at 2% as a safeguard, aimed at ensuring banks have enough capital to withstand rising vulnerabilities across the broader financial system.
In addition to market correction risks, the FPC identified cyber threats as the most immediate financial stability concern linked to frontier AI. The committee warned that AI technology could make cyberattacks on payment networks and banks faster, more sophisticated, and easier to scale. These risks are being monitored alongside geopolitical tensions, which the FPC notes are further complicating the global market environment.
While Bailey has previously acknowledged that AI could provide a “positive supply shock” to the economy through efficiency gains, the current focus has turned toward the “interconnected vulnerabilities” shared by technology giants and cloud providers. The concentration of AI infrastructure among a few large firms means that a failure or significant pullback in one area could have widespread consequences for the global financial architecture.
