AI cyber risk is biggest immediate threat to global financial system, warns Bank of England chief

Bank of England chief Andrew Bailey, who is also Financial Stability Board (FSB) Chair, made the comments in a letter to G20 finance ministers and central bank governors ahead of their meetings this week. Yui Mok/Pool via REUTERS(via REUTERS)AI Quick ReadBank of England Governor Andrew Bailey has warned that the impact of artificial intelligence on cyber risk is the most immediate concern facing the global financial system, saying advanced AI could dramatically alter the speed, scale and economics of cyberattacks.

Bailey, who is also Financial Stability Board (FSB) Chair, made the comments in a letter to G20 finance ministers and central bank governors ahead of their meetings this week.

Bailey said the rapid development of advanced AI models could allow cyber vulnerabilities to be identified more quickly, potentially giving attackers greater ability to exploit weaknesses across financial institutions and critical systems.

He warned that regulators and financial institutions could face significant operational and resilience challenges if their testing, patching and recovery processes fail to keep pace with the technology.

“Recent developments highlight the importance of ensuring that advances in capability are matched by resilience and preparedness,” Bailey said.

He called for safe and responsible AI model releases to be supported “on a global basis”.

The FSB chair also highlighted the financial sector's growing dependence on a relatively small number of powerful technology providers.

Such concentration, he warned, could pose risks beyond individual institutions and potentially undermine confidence across financial markets if a major technology provider experiences a disruption or cyberattack.

Bailey said many countries currently lack adequate systems to manage the deployment of increasingly advanced AI models, highlighting the need for regulators to strengthen preparedness.

Bailey's warning comes amid growing scrutiny of the ability of advanced AI systems to operate beyond their intended safeguards.

The concerns follow the US administration's tightly controlled rollout of Anthropic's powerful Mythos model, which at one point restricted access to US nationals.

In July, an OpenAI agent reportedly escaped a controlled testing environment and hacked AI company Hugging Face, raising further questions about whether increasingly capable AI systems could circumvent security safeguards.

Beyond AI-related cyber threats, Bailey reiterated his earlier concerns about vulnerabilities in global financial markets.

He pointed to stretched valuations linked to AI companies and weaknesses in government debt markets as potential sources of market corrections.

He also flagged rising use of leverage in equity markets as an emerging concern, warning that greater leverage could amplify losses during periods of market stress.