BoE article on frontier AI and the question of governance

The Bank of England (BoE) has published an insights article by Andrew Bailey, Governor of the BoE, on frontier AI and the question of governance. Mr Bailey argues that while the risks are real and increasingly significant, policymakers should not automatically turn to the questions of regulation. Instead, they should first ask what precisely the problem is that they are trying to solve. In his view, the challenge posed by frontier AI is that, in its most advanced forms, it threatens to operate outside society's framework of shared norms. A sufficiently powerful system functioning within a self-reinforcing loop risks reducing society's ability to exercise meaningful oversight and intervention. He is clear that this does not mean prohibiting frontier AI, whose potential benefits include driving scientific discovery, enhancing productivity and contributing to economic prosperity. However, society must retain the ability to intervene, to establish the boundaries within which these systems operate, and to revise those boundaries as the technology evolves.

Mr Bailey identifies rigorous model testing, conducted before and after deployment, as a sensible starting point. He notes that the UK's AI Security Institute is helping to develop the scientific and technical foundations of AI assurance, although the pace of progress must accelerate. In his view, testing should not be seen as an alternative to future regulation, nor as a complete solution. A more formal regulatory framework may well emerge over time, but understanding, testing, and establishing credible points of intervention must come first. Mr Bailey also links these issues to financial stability. He warns that frontier AI materially increases the scale and sophistication of cyber-threats facing the financial system, and that the resilience of payments networks, financial market infrastructures and banks can no longer be considered separately from AI advances. Rigorously testing the models that firms use will help to show how AI can be deployed safely. In time, that understanding could be codified into standards to ensure a consistent approach across the financial system.

Return to main website.