As enterprise AI adoption accelerates, security and governance controls are struggling to keep pace, leaving organizations exposed to new risks. Krista Case of theCUBE Research and Jon Oltsik, principal analyst in residence at SiliconANGLE and theCUBE, kick off day two of the AI Cybersecurity & Resilience Summit with a discussion on translating AI security frameworks into practical safeguards. They address gaps in policy enforcement, agent identity management, threat modeling and incident recovery, along with strategies for strengthening cyber resilience, maintaining human oversight and aligning AI risk management with business priorities.