Bank of England Warns AI Bubble Could Trigger Crisis

Andrew Bailey, who serves as the governor of the Bank of England and chair of the Financial Stability Board (FSB), raised the alarm in a letter to G20 finance ministers. He highlighted that while the global financial system has managed to absorb recent geopolitical shocks, it remains highly vulnerable. Bailey specifically pointed to the combination of inflated AI valuations and a surge in debt-fueled speculation as a dangerous mix that could amplify future market shocks.

The central banker expressed deep concern over the growing use of leverage among investors, including retail traders utilizing leveraged ETFs and hedge funds holding complex positions across equities and government bonds. This high concentration of debt, layered on top of stretched valuations, creates a fragile environment. Bailey warned that because of the dense web of cross-investments between AI startups and major cloud providers, a failure or downturn at a single prominent AI firm could quickly drag down other technology giants and spark a wider market collapse.

Beyond market stability, Bailey cautioned that frontier AI models present severe cyber security risks to the global banking infrastructure. He noted that these advanced models could "materially alter the speed, scale and economics of cyber risk," making attacks significantly cheaper and more frequent. Because the global financial sector relies heavily on a small group of dominant technology providers, a successful cyberattack on one vendor could simultaneously disrupt financial institutions across multiple nations.

To mitigate these threats, Bailey urged international policymakers to establish unified rules for the development and deployment of advanced AI, noting that many countries currently lack any regulatory framework for frontier models. The FSB is currently investigating how financial institutions can safely deploy frontier AI to bolster their cyber defenses. Bailey's warnings echo concerns from academic experts, such as NYU finance professor Aswath Damodaran, who recently cautioned that an AI market correction could be more damaging than the dot-com bust due to the massive, debt-financed physical infrastructure underlying the AI boom.