We have gone from AI that generates to AI that reasons to AI that builds and does things. We have gone from Generative AI to Agentic AI. While in theory and under closed conditions, an LLM is deterministic, in current practical applications it returns a probabilistic outcomes. At the same time, these highly probabilistic AI agents / agentic AI have increasingly become a part of business workflows, taking actions, making decisions and doing things for you. Agentic AI is making million dollar decisions while it still hallucinates, isn't always accurate, can be hacked, prompt injected, and doesn't have deterministic guardrails around it. In this keynote, I cover: 1) The Current State of Agentic AI,  2) Current Threat Landscape with AI 3) How to Govern and Secure Agentic AI

Looking to go from chaos and unpredictability to resilience in the world of AI? Start here with The Predictability Factor newsletter at The Monica Talks Cyber (https://www.monicatalkscyber.com).