Securing Agentic AI: Threats, Models, and Mitigation in the Age of Autonomy | Ryan Ray

Description

Explore real-world strategies for securing agentic AI systems. We'll frame enterprise AI agent adoption within governance models and shifting regulatory expectations, examining how agents reshape attack surfaces with novel threats like autonomous exploitation, goal manipulation, and AI-driven social engineering. By thinking like attackers, security professionals can incorporate adaptive trust models, secure-by-design principles, and AI-enhanced threat detection. We'll spotlight threat modeling techniques addressing unique risks including intent breaking and deceptive behaviors, providing a practical blueprint for securing intelligent systems. Subscribe to Zenity Labs for the most up-to-date research on the state and future of agentic security. https://labs.zenity.io/