Bank of England questions whether current oversight fits autonomous AI agents in finance.

In a Wharton School podcast, Bank of England Deputy Governor Sarah Breeden explains why agentic AI systems that execute trades or rebalance portfolios on their own challenge existing financial oversight. The Bank has tracked AI use in banks for five to seven years.

Sarah Breeden, deputy governor for financial stability at the Bank of England, tells the Wharton School podcast that her institution has tracked AI use in financial firms for five to seven years. Until now, the technology mainly supported fraud detection, cyber defense, or research tasks for staff. Agentic AI changes that picture: systems are given a goal and work out the steps themselves, whether executing financial transactions, rebalancing portfolios, or running cyber defenses, without a human deciding each individual action.

In Breeden's view, this shifts the ground under financial oversight. Existing rules assume that institutions make decisions through people; autonomous agents no longer fit that assumption cleanly. The accompanying Wharton piece cites a study on AI-powered trading, algorithmic collusion, and price efficiency, showing how automated systems can shape market behavior in new ways.

The conversation leaves open which benchmark will define success for new oversight rules and which concrete tools, such as mandatory kill switches or traceability of agent decisions, the Bank of England will eventually propose. For financial institutions running their own AI systems, this is a signal to review governance processes early, before binding requirements arrive.

This story was produced automatically from the source named above and checked by software before publication. The image is symbolic and shows neither the event nor a real person. How this paper is made