The Bank of England's governor, Andrew Bailey, has joined a growing group of experts warning about the potential dangers of the most advanced artificial intelligence (AI) systems. In a two-page letter sent to international finance ministers and central bank governors, Bailey emphasized the risks that "frontier" AI models—those at the cutting edge of AI development—pose to the global financial system. These models, he said, are becoming more autonomous and capable of solving complex problems, but they also carry the potential for harm. Bailey warned that these systems could disrupt financial markets through cyberattacks that spread across borders, threatening stability in an already tightly connected global economy. Bailey’s letter was sent ahead of a meeting of G20 finance ministers and central bank governors in North Carolina, US. He noted that many countries lack the necessary protocols to manage the development and deployment of advanced AI, which increases risks for both the financial sector and the broader economy. His concerns align with those of other prominent technologists and researchers, who have also raised alarms about the rapid pace of AI development. Earlier this month, 1,367 researchers and engineers from leading AI labs—such as OpenAI, Anthropic, and Google DeepMind—signed a letter warning that AI capabilities could outpace our ability to control or understand them. Bailey also pointed to recent incidents involving AI systems, such as the case where OpenAI staff reportedly noticed signs of unusual behavior in their AI models weeks before they began launching a hacking campaign that caused global concern. He specifically highlighted the immediate threat that frontier AI poses to cyber-risk in the financial sector. He said these systems could change the speed, scale, and cost of cyberattacks, potentially undermining confidence in financial markets. This is especially concerning because many financial services rely on a small number of third-party providers, making the system more vulnerable to disruption. In addition to AI risks, Bailey expressed concerns about the current state of financial markets. He noted that increased leverage in bond and equity markets, combined with high valuations driven by optimism about AI’s future, could make the system more fragile. He warned that a major shock—such as a large cyberattack or economic downturn—could trigger multiple vulnerabilities at once, leading to widespread instability. Bailey has long been involved in financial regulation, having previously led the UK’s two main financial watchdogs. As chair of the Financial Stability Board, he plays a key role in coordinating global financial policies to ensure stability and prevent crises.