Analysis
The report describes information-gathering, not new rules: sources said U.S. regulators are not yet being prescriptive. It lands alongside the IMF's warning on cyber risk and an April interagency model-risk guidance that expressly left generative and agentic AI out of its scope, with the agencies saying they plan to seek public input on those tools.
What Reuters reported
Banking regulators are stepping up efforts to contain growing AI risks, six people familiar with the matter told Reuters in an exclusive report published October 13, 2026. Some of the sources pointed to the potential for cyberattacks on banks or rogue AI agents to undermine global financial stability. Reuters said the details were being reported for the first time. [1]
Two of the people said three top U.S. regulators, including the Federal Reserve, are questioning banks about their use of advanced AI models. The questions concern whether banks use Chinese-developed technology, use AI agents in cyber defense and can control increasingly autonomous systems, according to three of the sources. [1]
What U.S. supervisors are asking
According to two of the people, questions from the Office of the Comptroller of the Currency (OCC), the Federal Deposit Insurance Corporation (FDIC) and the Fed focus on what limits are placed on AI systems, how their actions are monitored and approved, and whether banks can track and audit AI agent decisions. They also ask what safeguards stop agents from exceeding their authority, whether humans stay involved in key decisions, and how banks can shut down autonomous systems. [1]
The written and verbal inquiries go to large, mid-size and smaller banks, three of the people said, and regulators are still gathering information and are not yet being prescriptive about what lenders should do. The OCC, the Fed and the FDIC declined to comment. [1]
Europe
Three people with knowledge of the thinking of supervisors including the Bank of England and the European Central Bank said they are pressing banks to improve recovery plans, out of concern that AI advances could cause serious disruption in financial markets sooner than previously feared. Two of the people said British supervisors are encouraging the largest banks to examine "bare metal" recovery systems, which are entirely independent of a bank's main technology and can take over if primary systems are compromised; they said UK authorities intend to work with industry rather than introduce new requirements. The Bank of England declined to comment. [1]
Reuters noted that the ECB gave lenders an end-October deadline in July to produce plans against AI-enabled cyber threats, and that it pointed to that letter when asked about its strategy. [1]
Background and what is on the record
Reuters tied the scrutiny to recent incidents, including a statement from cybersecurity company CrowdStrike this month that a person in China likely used a Chinese-developed AI agent and Anthropic's Claude Code in a hack of South Korean banks, and to a call by JPMorgan Chase CEO Jamie Dimon for corporate leaders to join a U.S. industry group on AI risks. [1] These are Reuters' descriptions; the CrowdStrike finding is the company's own statement.
On the public record, the OCC, the Fed and the FDIC jointly issued revised model risk management guidance on April 17, 2026. It says generative AI and agentic AI models are "novel and rapidly evolving" and therefore outside its scope, and that the agencies plan to issue a request for information on banks' use of AI, including generative and agentic AI. [2] The IMF's October stability report, published the same day as the Reuters article, separately called for stronger cybersecurity capacity and tighter oversight of third-party technology vendors. [3]
What remains uncertain
The account rests on anonymous sources speaking to Reuters; no regulator confirmed the inquiries and the OCC, Fed, FDIC and Bank of England declined to comment. Apart from the article's republication by other platforms, no second outlet or official statement confirming the inquiries was found. What the inquiries may lead to is unknown.