Bank of England Governor warns that advanced AI could destabilize the highly interconnected global financial system via cyberattacks across jurisdictions
Andrew Bailey warns group about risk of advanced ‘frontier models’ destabilising the ‘highly interconnected’ financial system
Context & Ripple Effects
Bailey’s intervention extends an escalating supervisory focus: the ECB and ESRB had already called frontier models a systemic risk for lenders, while the ECB convened Eurozone banks to discuss exposure to the latest models and lessons from US institutions with Mythos access.
The concern is not merely an AI-market reversal. In 2023, Gary Gensler identified shared dependence on the same base models or data providers as a potential crisis channel; Bailey places cyber disruption and cross-border transmission at the center of that concentration risk.
First-order effects
- Bank of England Governor Andrew Bailey puts frontier-model cyber risk before the G20, directing attention to how an attack or failure could damage confidence across connected financial markets.
- Banks and financial authorities face a sharper operational-risk question: whether AI systems and their dependencies can be secured and contained across jurisdictions during a cyber incident.
Second-order effects
- The warning increases pressure on lenders to scrutinize concentrated model, data and infrastructure dependencies rather than assessing AI tools only as isolated vendor risks.
- Cross-border supervisors have greater reason to coordinate incident-response expectations, because a cyber event affecting common AI services can propagate beyond any single bank or national regulator.
Third-order effects
- If institutions continue adopting a narrow set of frontier models and shared providers, financial-stability oversight is likely to treat AI concentration as a system-level operational vulnerability alongside firm-level cyber controls.
- The policy challenge shifts toward governance that can test and manage AI-linked contagion across national boundaries, not solely within individual institutions.
The trend: Financial regulators are moving from broad AI-risk warnings toward treating shared frontier-model dependencies and AI-enabled cyber disruption as potential channels of systemic contagion.