A Reference Architecture for Governing Autonomous, Non-Human Identities in the Regulated Enterprise

Author(s)	Madhusudan Kousik Kappaganthula
Country	United States
Abstract	Agentic AI introduces a new class of non-human identity—autonomous, fast, delegated, and rapidly proliferating—that traditional identity and access management was never designed to govern. Left unmanaged, agents become over-permissioned, unowned, and capable of causing harm at machine speed. This paper presents a vendor-neutral reference architecture that treats agent identity as a first-class category, governed by an agentic identity control plane spanning discovery, authentication, mandate-based delegation, continuous authorization, just-in-time access, lifecycle governance, and closed-loop observability. Anchored on open standards and zero standing privilege, it produces a regulator-ready audit trail correlating identity, delegation, action, decision, and outcome.
Keywords	Agentic AI; AI agents; non-human identity (NHI); identity and access management (IAM); agent identity governance; zero standing privilege; just-in-time access; on-behalf-of delegation; mandate-based delegation; continuous authorization; workload identity; zero trust; identity governance and administration (IGA); secrets management; policy-as-code; observability and audit; behavioral risk scoring; multi-cloud agent-to-agent (A2A); machine identity security.
Field	Computer > Network / Security
Published In	Volume 7, Issue 8, August 2026
Published On	2026-08-25
DOI	https://doi.org/10.70528/IJLRP.v7.i8.2322

About IJLRP Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics	Join as a Reviewer Editors & Reviewers Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijlrp.com

International Journal of Leading Research Publication