All papers  /  No. 06  /  AI Agent Governance

The Alignment Tax Governing AI Agent Identity Inside the Enterprise IAM Estate

Prabin ShresthaOctober 2026Research preprint

Abstract

Autonomous AI agents now run in production as identity subjects that enterprise identity and access management (IAM) was never built to govern. An agent authenticates programmatically, executes at machine speed, delegates authority to sub-agents, crosses organizational trust boundaries, and persists after the business purpose for its privileges has expired. The governance gap has three precise parts: no authoritative lifecycle source defines when an agent identity begins or ends; ownership and attestation for non-human identities remain unclear in most organizations; and no single standard states how agent identities are issued, scoped, and retired inside the enterprise IAM estate. The paper names the recurring cost of closing that gap the alignment tax and answers with an agent identity control plane that extends existing IAM practice to agentic workloads. The analysis builds on the Cloud Security Alliance’s Agent Identity Governance Framework (2026), the NIST AI Risk Management Framework and its Generative AI Profile, and the OWASP Top 10 for LLM Applications (2025), together with recent platform moves by SailPoint. Contributions are (1) a definition of the alignment tax in five measurable dimensions (lifecycle, access-modeling, attestation, credential, and evidence); (2) a mapping of what the existing standards and frameworks cover and what each one leaves to the IAM team; and (3) an agent identity control plane, with practitioner steps for extending a platform such as SailPoint Identity Security Cloud to agentic workloads.

AI agent identityIAMidentity governancenon-human identityNIST AI RMFSPIFFESailPoint

Download & deposit

© 2026 Prabin Shrestha. Licensed under CC BY-ND 4.0: share unchanged with attribution; no derivatives.

Cite this paper

Shrestha, P. (2026). The Alignment Tax Governing AI Agent Identity Inside the Enterprise IAM Estate. Research preprint. https://prabin-iam-ai-papers.pages.dev/papers/alignment-tax.html
@misc{shrestha2026alignmenttax,
  author = {Shrestha, Prabin},
  title = {The Alignment Tax Governing AI Agent Identity Inside the Enterprise IAM Estate},
  year = {2026},
  month = {oct},
  url = {https://prabin-iam-ai-papers.pages.dev/papers/alignment-tax.html},
  note = {Research preprint, CC BY-ND 4.0}
}
← Previous paperGuests in the Machine: Extending Third-Party Identity Governance to Vendor-Embedded AI Agents

Keep reading

Related papers