The Bank of England is reviewing its regulations regarding agentic AI in finance, including payments, trading, cybersecurity, and operations, as stated by Deputy Governor Sarah Breeden during the European Central Bank Forum in Portugal. Breeden emphasized that existing frameworks were not designed for AI systems that can operate autonomously without direct human instruction.
Understanding Agentic AI in Financial Workflows
Agentic AI refers to systems capable of making decisions and executing tasks independently. In the finance sector, these systems are increasingly utilized for various applications, such as product recommendations and trading operations. Unlike traditional automated trading tools, agentic AI can pursue specific objectives with minimal human oversight.
Breeden noted that these systems could behave similarly if trained on comparable data or designed with aligned goals. A 2026 report from the Cambridge Centre for Alternative Finance revealed that 81% of financial firms are adopting AI technologies, with 52% actively integrating agentic AI into their operations. However, most applications remain focused on internal functions like process automation and knowledge management.
Cyber Resilience Risks Associated with Agentic AI
Breeden identified cyber resilience as a significant concern regarding the use of agentic AI within financial systems. She described the current technological advancements as a “step change” in cyber capabilities, urging supervisors to evaluate risks across the entire financial system rather than on an individual firm basis. While AI tools can enhance cyber defenses, there is a risk that malicious actors could exploit these same tools to destabilize financial systems.
