LONDON, June 30 - The rapid development of autonomous artificial intelligence agents has highlighted potential weaknesses in how financial stability is overseen, a senior Bank of England official said on Tuesday.
Speaking at the European Central Bank Forum on central banking in Portugal, Sarah Breeden, the Bank's deputy governor for financial stability, said current supervisory systems did not anticipate agents that can act with a high level of autonomy. She stressed that relying on a human to be involved in every decision an agent makes is unlikely to be feasible.
"Our frameworks were not built to contemplate autonomous agents, and relying on a human in the loop for all agent actions is unlikely to be realistic. More sophisticated governance and accountability frameworks may be needed," Breeden said.
Her remarks come amid broader concern among regulators and international standard-setting bodies about how AI is being deployed across the financial sector. The discussion around agentic AI intensified following the release of Mythos by Anthropic, which analysts say could introduce meaningful cybersecurity challenges for banks and other financial institutions.
Earlier in June, the Financial Stability Board urged the adoption of tighter safeguards to address the specific risks posed by AI agents. The FSB highlighted that these agents present a distinct challenge to traditional models of human oversight.
Breeden's comments reflect an acknowledgment at high levels of central banking that governance and accountability arrangements may need to evolve in step with the technology. While she did not outline specific policy prescriptions, the emphasis was on adapting frameworks to cover autonomous activity rather than assuming continued human supervision at all times.
The remarks add to an ongoing policy conversation about how to monitor and contain risks from new AI capabilities within financial systems. They suggest regulators may need to consider more detailed rules or standards that specifically address agentic behavior and the operational and cybersecurity vulnerabilities linked to such systems.
Summary
Sarah Breeden of the Bank of England warned at the ECB Forum that current regulatory frameworks may be ill-suited to autonomous AI agents, and said more advanced governance and accountability frameworks might be necessary. Regulators and the Financial Stability Board have voiced similar concerns, especially following the release of Anthropic's Mythos, which analysts say could create cybersecurity risks for the banking industry.