LLOYDS Banking Group is advancing its approach to agentic AI security by treating it as an engineering challenge rather than a theoretical concept. During a session at the OWASP GenAI Security Summit, Manija Poulatova and Kirsty Montignani detailed the bank's proactive strategy for integrating AI security across product lifecycles. The strategy centers on 11 AI use cases with security as the 12th focus, aiming for low-risk, high-value applications.
Key initiatives include creating an internal agent marketplace for better governance and identity management, especially in defining agent identities distinct from human ones. Lloyds is employing automated testing for red-teaming exercises to scale defensive measures. The bank's successful AI implementations promise to deliver significant value, enhancing customer service and operational efficiency.