Andrew Bailey warns G20 of danger AI poses to financial system

The threat that frontier AI models pose to cyber security is increasing the risk of a market meltdown, the head of the world’s financial stability watchdog has warned, calling for more countries to introduce controls on the technology.

In a letter to G20 finance ministers and central bank governors, Andrew Bailey said financial services companies and technology groups must prepare for “more severe scenarios involving simultaneous disruption across multiple firms or shared technology dependencies”.

The letter from the Bank of England governor, who also chairs the Financial Stability Board, underlines how the guardians of the world’s financial system are becoming increasingly alarmed about the dangers of AI-enabled cyber attacks.

These fears intensified following a series of recent incidents in which the flagship models being tested by Anthropic and OpenAI went rogue, hacking into external organisations and creating fake identities to deceive people who were running the tests.

Bailey said in his letter, published on Monday, that AI cyber risks were adding to existing vulnerabilities in the financial system, including energy-driven inflationary pressures, rising interest rates, increased investor leverage and stretched equity valuations.

Urging more countries to take “appropriate steps” to control the release of new frontier AI models, Bailey appeared to be trying to convince the US to reconsider its hands-off approach to regulating the new technology.

“Recent developments have also highlighted to me that many jurisdictions do not have the protocols in place to manage the development, release and deployment of advanced frontier AI models, heightening risks for the financial sector and beyond,” he wrote.

US President Donald Trump recently signed an executive order that stopped short of giving the government power to block the release of an AI model, even as it set up a voluntary framework for various agencies to inspect frontier models ahead of their rollout.

Earlier this year the White House imposed export restrictions on Anthropic’s frontier AI models over concerns they could be used to exploit cyber security vulnerabilities in key infrastructure, but these were removed in July after the company agreed to extra safeguards.

When Chinese President Xi Jinping visits Washington for a meeting with Trump on September 24, they are expected to discuss co-operation on AI regulation between the two countries that are home to most of the main developers of the technology.

Bailey called for more financial groups to develop so-called “bare metal” backup systems that are kept completely unconnected from their main networks in order to be able to restore IT operations after a major cyber attack.

The FSB is examining how AI models can be safely deployed by financial firms and how they can improve their resilience to cyber security attacks, he added.

Finance ministers and central bank bosses of G20 countries are meeting in Asheville, North Carolina, on Monday and Tuesday.

The FSB is a global watchdog that brings together finance ministry officials, central bankers and securities regulators from G20 countries. It includes officials from the US, UK, Canada, France, Germany, Japan, Saudi Arabia, Australia and China.

Listing likely threats to the financial system, Bailey said: “Markets remain vulnerable to a potentially disorderly correction that could spread across borders.”

He said “fragilities in sovereign debt markets” were among his concerns, as well as signs of trouble in private credit markets, and stretched valuations in equity markets fuelled by optimism about the potential of AI.

Growing use of debt, or leverage, by investors in equity markets also makes a crisis more likely, Bailey said. “As we have seen multiple times in the past, rising leverage is a feature of a maturing financial cycle,” he said.