AI agents introduce a new set of identity and access challenges. They can act independently, on behalf of users, and across multiple systems, making traditional approaches based on standing privileges and static credentials difficult to manage and audit.

Join us for a hands-on technical workshop where you’ll work directly with HashiCorp Vault in an Amazon EKS environment to explore how organizations can secure agentic applications using identity-based access, just-in-time credentials, least-privilege controls, and auditable authorization.

Through guided exercises, you’ll build and inspect progressively more sophisticated AI-agent scenarios, from workload identity and read-only access to personalized access and privileged actions requiring human approval.

You’ll leave with practical experience applying these security controls on AWS and a clearer understanding of how HashiCorp Vault can help secure AI-agent access to sensitive systems and data.

Using a working reference implementation, you’ll get hands-on experience with approaches designed to:

Establish verifiable agent identity across AI-driven workflows. Reduce standing privileges by using just-in-time Vault credentials. Connect user intent to agent actions as requests move across systems. Enforce access controls at the point of use, including database-level access. Create correlated audit evidence across user, workload, and data environments.

Hands-on workshop This isn't just a technical overview. Attendees will spend 90 minutes working hands-on with HashiCorp Vault, applying core Vault capabilities in a guided AWS-focused environment.

Workshop audience This Level 300 workshop is designed for experienced technical practitioners who already have familiarity with HashiCorp Vault, cloud security, identity, or secrets management.

• Cloud and AWS architects
• Cloud security architects and engineers
• DevSecOps and platform engineering teams
• Identity and access management (IAM) practitioners
• Secrets management and infrastructure security teams
• AI/ML platform and security teams
• Security leaders evaluating approaches to securing AI agents