The rise of AI agents represents a significant evolution in managing non-human identities, challenging IAM teams to rethink traditional patterns around service accounts, delegation, and trust boundaries.

When dozens or hundreds of agents make decisions and take actions on behalf of users, fundamental questions arise: Should they use service accounts? User credentials? A hybrid of both?

Drawing from real-world implementations, we'll explore

• Why agents strain traditional identity models and service account patterns
• Key decisions in implementing agent identity [with examples of what works/fails]
• How to evaluate risk when agents act on behalf of users
• Threat modeling unique to agentic systems
• Practical approaches to authentication and authorization for agents

You'll leave understanding:

• A clear understanding of the identity challenges agents create and how IAM teams can address them.
• Actionable strategies for evaluating risk and designing authentication and authorization for agents.
• Practical insights into where existing identity patterns fall short and how to evolve them for the AI era.