US regulators increase scrutiny on AI risks in banking sector

CrowdStrike: CrowdStrike is a major US-based cybersecurity company that provides threat intelligence and protection services. It reported that a person in China likely used Chinese-developed AI and Anthropic's Claude Code in a hack targeting South Korean banks. Jamie Dimon: Jamie Dimon is the chairman and chief executive officer of JPMorgan Chase, a leading global bank. He has urged corporate leaders to join a US-focused industry group to collectively address emerging AI risks following recent incidents. US regulators: US banking regulators encompass agencies such as the Office of the Comptroller of the Currency, the Federal Deposit Insurance Corporation, and the Federal Reserve that oversee financial institutions. In this news, they are actively quizzing banks on their deployment of advanced and agentic AI models, including questions about foreign technology, autonomous systems, and cyber defenses. Bank of England: The Bank of England is the central bank of the United Kingdom that supervises financial stability and banking operations. It is among the European supervisors urging major banks to enhance bare metal recovery plans and address potential AI-driven disruptions in financial markets. Federal Reserve: The Federal Reserve is the central banking system of the United States responsible for monetary policy and financial supervision. It is one of the three top US regulators directly inquiring about banks' use of frontier AI, agentic systems, and safeguards against rogue agents or cyber threats. International Monetary Fund: The International Monetary Fund is an international organization that promotes global financial stability and economic cooperation. It recently warned against complacency in markets and highlighted the risk that a major cyberattack on a critical financial institution could undermine confidence. Regulation: US regulators including the Federal Reserve, OCC, and FDIC are gathering information from banks on limits, monitoring, and human oversight of agentic AI systems without yet issuing prescriptive rules. Cybersecurity: European supervisors are encouraging banks to review independent bare metal recovery plans and use frontier AI for vulnerability identification while noting significant gaps in some firms' preparedness. Industry Response: Banks are competing to adopt agentic AI for tasks with minimal supervision while facing challenges such as IT weaknesses and dependencies on external providers flagged in recent incidents.