Bank of England Governor Andrew Bailey has stated that formal regulation of artificial intelligence is “not the right place to start” for managing the technology’s impact on the financial sector. In a position published on September 30, 2026, Bailey argued that the immediate priority should be the development of technical safeguards, such as “kill switches,” and rigorous testing protocols rather than new legislative statutes.
The Governor’s stance comes as the Bank of England monitors a significant surge in AI-related financial activity. According to the Financial Policy Committee (FPC) record for September 2026, AI-related debt issuance reached $450 billion in the first nine months of the year. This figure now exceeds the total value of gilts—approximately $333 billion (£339 billion)—that the UK government is expected to issue for the entirety of 2026.
Prioritizing Technical Intervention
Bailey’s “safety-first” hierarchy suggests that the speed of AI development requires “credible points of intervention” that can be deployed at a technical level. The concern centers on “frontier AI,” where recursive learning might create a “closed loop” in which models autonomously refine their own reasoning and decision-making processes without human oversight.
To prevent such systems from creating systemic instability, Bailey is calling for technical tools that allow human operators to halt or reset autonomous processes before they can trigger cascading market errors. This approach favors operational resilience over the often slower process of drafting and implementing international law.
Market Concentration and Valuation Risks
The scale of capital flowing into the sector has raised alarms regarding market stability. The Bank of England warned in its recent assessment of a potential “sharp market correction” in AI valuations. This follows a volatile period in July 2026 where high-growth tech stocks faced significant pressure.
The FPC indicated that if the anticipated productivity gains from AI do not materialize at a rate that justifies the current $450 billion debt pile, the sector could face a deeper correction than previously seen. The Financial Stability Board (FSB) has similarly flagged risks regarding market concentration, as a small number of dominant AI providers become deeply integrated into the infrastructure of global finance.
Global Context and “Morally Binding” Agreements
While the UK central bank moves toward a technical-first framework, other international efforts are focused on voluntary commitments. At a recent summit in the United States, major AI developers including OpenAI and Anthropic signed safety documents described as “morally binding.”
However, Bailey’s emphasis on “rigorous” testing suggests that the Bank of England views voluntary agreements as insufficient for the high-stakes environment of central banking and national financial stability. By focusing on kill switches and intervention points, the BoE aims to ensure that even if legislative frameworks remain in flux, the physical and digital ability to control AI-driven market moves remains with human regulators.