AI agents are becoming an integral part of electronic communications, connecting with other agents, MCP servers, microservices, and an expanding ecosystem of connected services and devices. As these interactions grow in scale and autonomy, security and trust will be critical to building a robust environment for agentic AI.

This keynote will examine the security models that can strengthen trust across AI systems and agent communications, exploring the advantages, limitations, and tradeoffs of different approaches. Drawing comparisons with other globally connected systems, including e-commerce and payments, Taher Elgamal will explore lessons from the evolution of digital security and what they can teach us about securing the emerging agentic AI ecosystem.