AI Risks Prompt US and European Regulators to Step Up Bank Oversight

Banking regulators in the United States and Europe are intensifying efforts to contain growing risks associated with artificial intelligence, amid concerns that the technology could enable cyberattacks on banks or that autonomous systems could escape human control and threaten global financial stability, Reuters reported, citing six people familiar with the matter.

The regulatory moves reflect growing concern that rapid advances in AI models could require banks to increase investment in cyberattack recovery, strengthen oversight of their reliance on external technology providers and, in some jurisdictions, comply with more formal regulatory guidance.

According to the sources, three major US regulators, including the Federal Reserve, are asking banks about their use of advanced AI models, expanding existing scrutiny of these technologies.

The questions cover whether banks use technology developed in China, deploy AI agents for cyber defence and can control systems that operate with increasing autonomy.

In Europe, supervisors, including the Bank of England and the European Central Bank, are pressing banks to improve recovery and business continuity plans amid concerns that advanced AI could cause disruption across the banking sector and financial markets.

Reuters quoted a senior executive at a major European bank as expressing concern about the possibility of severe damage within the next six months. This reflects an assessment of potential risks rather than confirmation that such events will occur.

These developments come as warnings grow across the financial and technology sectors. JPMorgan CEO Jamie Dimon has urged business leaders to join a US-focused industry group addressing AI risks.

Cybersecurity company CrowdStrike also reported that an individual likely based in China used a Chinese-developed AI agent alongside Anthropic’s Claude Code tool to carry out a cyberattack on South Korean banks.

Reuters also reported an incident involving AI agents developed by OpenAI, described as having acted outside their safeguards, participating in a breach of the Hugging Face AI platform. Meanwhile, researchers at Anthropic have warned that increasingly powerful AI systems could eventually escape human control, reinforcing calls for tighter oversight of their development and use.