Banking watchdogs ramp up AI scrutiny as stability fears mount

Banking regulators are intensifying efforts to address growing concerns over artificial intelligence (AI) risks, according to sources. Some experts warn that AI advances could lead to cyberattacks on banks and undermine global financial stability. Regulators in the US, including the Federal Reserve, are questioning banks about their use of advanced AI models, focusing on areas such as AI agents in cyber defense and the use of Chinese-developed technology.

In Europe, regulators like the Bank of England and the European Central Bank are urging banks to enhance their recovery plans due to fears of potential harm and disruption in financial markets caused by rapid AI advancements. The debate over AI oversight has been fueled by recent security incidents, with JPMorgan CEO Jamie Dimon advocating for a US-focused industry group to tackle AI risks.

US cybersecurity firm CrowdStrike reported a recent Chinese-developed AI agent used in a hack on South Korean banks. Additionally, OpenAI's Claude Code was implicated in a breach of AI platform Hugging Face. Researchers from Anthropic have expressed concerns about powerful AI systems potentially escaping human control, likening them to an existential threat.

The International Monetary Fund has also cautioned against complacency in the face of various risks, including cyberattacks. Banks are racing to adopt agentic AI, which can perform tasks with minimal human supervision, while grappling with the challenges of new models, such as IT weaknesses. Regulators are probing specific issues tied to advanced AI models, including autonomous AI agents, foreign-developed models, and new cyber risks.

In the UK, supervisors are encouraging large banks to examine "bare metal recovery plans," independent systems that can take over if primary operating systems are compromised, despite the high costs associated with building and maintaining such advanced recovery systems.