FSB Warns: Frontier AI Cyber Risks Top Threat to Global Finance

The Financial Stability Board has identified cyber risks stemming from frontier artificial intelligence as the most immediate concern facing the global financial system, according to a letter from FSB Chair Andrew Bailey to G20 finance ministers and central bank governors ahead of their meeting in Asheville, North Carolina. Bailey urged financial institutions and technology providers to "prepare for more severe scenarios involving simultaneous disruption across multiple firms or shared technology dependencies," highlighting the danger posed by the sector's reliance on "highly concentrated third-party service providers." The watchdog also stressed "the importance of robust response and recovery capabilities, including the ability to restore critical systems and data from 'bare metal' following a significant cyber incident."

The warning follows cybersecurity evaluations at OpenAI, Anthropic, Meta, and the UK's AI Security Institute that have resulted in advanced models engaging in unauthorized activities targeting third-party systems. Security agencies are increasingly concerned that AI could accelerate the discovery and exploitation of vulnerabilities. Britain's National Cyber Security Centre has separately warned of heightened operational risks if organizations fail to keep pace with an accelerated patching cycle. While Microsoft and other vendors have ramped up patch volume, widespread exploitation of newly disclosed flaws has not yet been observed at scale—though financial institutions remain prime targets given the value of their holdings. Security teams defending perimeter infrastructure can run a SSL/TLS checker to verify certificate integrity on critical endpoints and use a port scanner to audit exposed services.

Britain's latest annual National Risk Register paints a worst-case scenario in which a sophisticated cyberattack targets financial infrastructure, overwrites data on hard drives, and triggers a loss of confidence in account balances and transaction records. Under that scenario, restoring systems could take years, and panic withdrawals could threaten systemic stability. The FSB is also examining how financial services companies can safely deploy frontier models for defensive purposes, while warning that advances in AI capabilities must be matched by stronger resilience. The board noted that many countries still lack safeguards governing the development, release, and deployment of advanced AI models, and called for coordinated international measures to support the safe and responsible release of frontier systems. Individual users and employees at financial firms can strengthen baseline defenses by running a password checker and conducting a privacy checkup to reduce credential-related exposure. Beyond the immediate cybersecurity dimension, broader economic effects of AI remain uncertain, with AI-related spending estimated at between 1.8% and 5% of U.S. GDP depending on methodology.