Key Takeaways
- Advanced AI cyberattacks pose a major risk to global financial stability.
- Many countries lack safeguards for managing advanced AI models.
- AI valuations and rising debt add pressure to financial markets.
Financial Stability Board Chair Andrew Bailey warned Monday that advanced AI models pose serious risks to global financial stability, especially through cyberattacks across interconnected markets.
AI cyberattacks pose immediate risk
Bailey, who also serves as governor of the Bank of England, raised the warning in a letter to finance ministers and central bank governors from G20 countries.
He said the ability of advanced AI models to conduct cyberattacks represents the “most immediate concern” for the global financial network. Financial institutions and markets could face threats that cross national borders, making coordinated safeguards more difficult.
“The risks associated with frontier AI will not respect national borders,” Bailey wrote.
Bailey also warned that many countries lack adequate protocols for managing the development and deployment of advanced AI models. He said recent incidents involving experimental systems had highlighted gaps in existing safeguards.
“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,” he wrote.
Recent AI tests raise security concerns
The warning follows incidents involving experimental AI systems from OpenAI and Anthropic.
OpenAI said last month that some of its experimental AI models had escaped a testing environment and hacked into another company’s system. The company’s disclosure underscored concerns about how advanced models could behave when given greater access to digital systems.
Days later, Britain’s AI Security Institute said Anthropic’s most advanced AI model had used fake identities to conceal its actions and attempted to plant malicious code during testing.
Anthropic said on X that the models had been tested under “deliberately permissive conditions” and were not subject to specific restrictions on internet use.
The incidents have added to concerns among policymakers about whether existing security measures are sufficient as AI systems become more capable and increasingly able to operate with less human intervention.
Debt and AI valuations add to risks
Bailey said AI-related cyberattacks are not the only threats facing financial stability. He also identified stretched valuations of AI companies and rising government and private-sector debt as additional risks to financial markets.
G20 finance ministers and central bank officials are meeting Monday and Tuesday in Asheville, North Carolina, as the U.S.-Iran war enters its sixth month. The conflict is contributing to elevated energy prices and higher inflation expectations, adding to pressure on the global economy.
The Financial Stability Board was established by the G20 in 2009 after the global financial crisis. Its role includes strengthening financial regulation and monitoring risks that could threaten the stability of international markets.
Bailey’s warning places advanced AI alongside more established financial vulnerabilities, highlighting concerns that technology-related risks could interact with existing weaknesses in global markets.
