Andrew Bailey, governor of the Bank of England, warned the G20 that developments in frontier AI could pose a threat to global financial stability.
The Bank of England, which oversees monetary policy and financial-stability oversight in the UK, frames the concern around AI’s potential to change the speed, scale and economics of cyber risk and to undermine market confidence.
Shift of tone
That warning completes a shift in tone from earlier in 2026, when policymakers judged adoption had not yet become systemically risky, to July and August moves that treat rapid frontier-AI progress as an emergent systemic threat.
The Bank highlights four transmission channels where AI could amplify risk: use of advanced models in firms’ core decisions, effects on market behaviour, concentration and dependencies on a small set of AI providers, and a faster, more capable cyber-threat environment.
Bespoke rule?
Regulators have already signalled that more capable, agentic AI may require bespoke rules to protect operational resilience and third-party risk, and the Bank flagged the possibility of a sharp market correction if AI-linked expectations reprice.
Bailey’s intervention asks G20 peers to prioritise cross-border coordination on cyber resilience and third-party concentration as AI capabilities continue to scale.
