AI Identity Security Engineering

By: John, Charles

AI agents can act across tools, databases, cloud services, and enterprise systems. Each connection raises a security question: whose authority is the agent using, what is it allowed to do, and how can that authority be withdrawn?

AI Identity Security Engineering gives experienced engineers a framework for securing the identities, credentials, sessions, and permissions behind AI systems. It follows authority from the user and agent through tokens, gateways, connected tools, retrieval systems, and protected resources.

The book covers workload and agent identity, OAuth and OpenID Connect, token exchange, delegated access, capability-based authorization, short-lived credentials, tenant isolation, and policy enforcement. It also examines prompt injection as an authorization problem, credential theft, lateral movement, detection, revocation, incident response, and recovery.

Architecture reviews, engineering patterns, testing methods, and reference designs connect these concepts to production decisions. The focus throughout is on making an agent's authority specific, observable, limited in time, and recoverable when trust fails.

Written for security engineers, IAM architects, platform engineers, software architects, and technical leaders building or governing AI systems.

• Format: Paperback
• Language: English
• ISBN: B0HKTK8VXX
• ISBN13: 9798185870686
• Release Date: September 2026
• Publisher: Independently Published
• Length: 558 Pages
• Weight: 2.41 lbs.
• Dimensions: 1.1" x 8.0" x 10.0"