Zero Trust Authorization for AI Workloads

You'll walk away able to: build zero-trust authorization for AI agents, end to end.

Register NowPresented byDhaval Desai

• ✓Enrol agents with dynamic client registration and chain identity from human to agent with on-behalf-of tokens
• ✓Express authorization as policy-as-code and run an embeddable policy decision point
• ✓Attach authorization in an agent gateway and protect MCP tools
• ✓Protect a Spring Boot backend with a separate PDP service over the AuthZEN APIs

Participants will build an end-to-end, zero-trust security infrastructure for AI Agents.

Participants will learn:

How to use dynamic client registration to enroll agents

How to use `on-behalf-of` tokens to create a chain of identity from human-to-agents

Use policy-as-code paradigm for modern authorization

Deploy embeddable policy decision point to implement zero trust authorization at different layers of agentic application

How to deploy and attach authorization in agent gateway

How to protect MCP tools using agent authorization

How to deploy separate PDP service that uses AuthZEN APIs to protect backend spring-boot java service

How participants will build it:

Participants will install docker based components and configure them to connect with each other.

Participants will build a small spring-boot java app that exposes APIs. Some of these will support MCP tools.

For reference, a step-by-step README will be provided.

Dhaval Desai

Community Manager at Gluu Inc.

Dhaval is Community Manager at Gluu Inc., where he is currently building open-source identity, access, and governance software for agentic systems.