(TR03) (Secure AI) A discussion on AI security in 2026, including agentic systems, identity, and red teaming

As autonomous AI agents move into production, the attack surface has fundamentally shifted. This session breaks down the three critical shifts security teams must master: vulnerabilities in agentic protocols like MCP, the failure of traditional identity when AI acts on behalf of humans, and the evolution of red teaming for non-deterministic, multi-step systems. Join CoSAI for a deep dive into the practical frameworks and real-world threat models defining the next era of secure AI.