Securing Agents with Open Standards

Learn how to implement production-ready security for AI agents through this 19-minute conference talk from the AI Engineer World's Fair. Explore the critical identity and authorization challenges that emerge when moving AI agents from prototype to production, including ensuring agents act only for authorized users, preventing over-broad API access and data leaks, handling user approvals in headless environments, and avoiding common pitfalls like manual credential sharing and stale keys. Discover the technical trade-offs behind building secure, user-aware AI agents as the speakers dive beyond surface-level recommendations to explain architectural decisions and integration strategies using OAuth, OIDC, RAR, and asynchronous authorization protocols. Watch a hands-on demonstration featuring an open-source Node.js agent implementation that showcases practical integration patterns without relying on proprietary solutions, while gaining insights into how these security approaches are shaping the future of identity management in generative AI applications.