A critical issue is emerging as organizations embed AI agents into workflows: securing agentic AI requires a very different approach to security. In this episode, Accenture CISO Kris Burkhardt breaks down what's at stake with Daniel Kendzior who leads Security AI Reinvention for Accenture Cybersecurity and Wei Liu who leads Cyber Strategy, Risk, and Innovation for Accenture Information Security. Discover why traditional security frameworks fall short for agentic systems, how to create security boundaries without stifling progress, and why starting small and scaling fast is a strong strategy. The big takeaway? Security practitioners can't let the excitement around agentic AI distract them from security basics.