An acceptable use policy defines which data, tools, workflows, and actions are permitted for an identity or system. For AI governance, it becomes the boundary that turns vague intent into enforceable scope, which auditors and security teams can test against actual runtime behaviour.
Expanded Definition
An acceptable use policy, or AUP, sets the operational boundary for what an identity, application, or AI agent may do, what data it may touch, and which tools or workflows are permitted. In NHI governance, the policy is most useful when it is specific enough to be enforced at runtime and reviewed after the fact. That makes it different from a high-level ethics statement or a broad security charter, which may describe intent but not actual guardrails.
Definitions vary across vendors when AUPs are applied to agentic systems, because some teams treat them as user conduct rules while others use them as machine-enforceable scope controls. For NHI Management Group, the practical meaning is narrower: the AUP should translate approved purpose into testable constraints that security, audit, and platform teams can validate. That often includes allowed APIs, approved data classes, outbound destinations, escalation conditions, and prohibited autonomous actions, consistent with governance expectations in the NIST Cybersecurity Framework 2.0. The most common misapplication is treating the AUP as a policy document that exists only for compliance review, which occurs when no one maps it to controls, logs, or enforcement points.
Examples and Use Cases
Implementing an AUP rigorously often introduces friction for developers and platform engineers, requiring organisations to weigh flexibility against the cost of tighter control and monitoring.
- An internal AI coding assistant is allowed to read approved repositories but blocked from exporting secrets, matching the lifecycle control focus described in Ultimate Guide to NHIs — Lifecycle Processes for Managing NHIs.
- A service account may call ticketing and logging APIs, but not finance systems, unless a change ticket changes the approved scope.
- A chatbot can summarize customer records, yet it is prohibited from writing back to production systems or sending data to unapproved third-party endpoints.
- A CI/CD automation identity is permitted to deploy only signed artifacts from a known pipeline, while access to admin consoles is excluded by policy.
- An AI agent used for procurement can draft purchase requests, but final submission requires human approval because autonomous execution is outside the approved use case.
Good AUP design also needs auditability. That is why NHI teams often pair it with asset and secret inventories, then validate deviations against the controls discussed in Ultimate Guide to NHIs — Regulatory and Audit Perspectives and policy alignment guidance from CISA Zero Trust Maturity Model.
Why It Matters in NHI Security
An AUP is one of the few documents that can turn a broad trust relationship into a measurable boundary for NHI activity. Without it, service accounts and agents tend to inherit implicit permissions, which is how over-privilege, secret misuse, and unsafe tool access spread. The issue is not theoretical: NHI Mgmt Group reports that 96% of organisations store secrets outside of secrets managers in vulnerable locations, making unclear use boundaries even more dangerous when credentials are reused across tools and environments.
AUPs matter because they give security teams something to test against during access reviews, incident response, and agent onboarding. They also support zero trust by making permission scope explicit rather than assumed, which aligns with NIST Cybersecurity Framework 2.0 and common audit expectations for least privilege. Organisations typically encounter the need for an AUP only after an agent has sent data to the wrong system, overreached its intended workflow, or consumed privileged credentials in an unapproved way, at which point the policy becomes operationally unavoidable to address.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
OWASP Non-Human Identity Top 10 and OWASP Agentic AI Top 10 address the attack and risk surface, while NIST CSF 2.0 set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| OWASP Non-Human Identity Top 10 | NHI-01 | AUPs define what an NHI may do, which maps to permission scope and misuse prevention. |
| NIST CSF 2.0 | PR.AA-5 | AUPs support identity governance by limiting access to authorised functions and assets. |
| OWASP Agentic AI Top 10 | A1 | Agentic systems need bounded tool and data use to prevent unsafe autonomous behaviour. |
Constrain each NHI to approved actions, tools, and data paths, then verify runtime enforcement.
Related resources from NHI Mgmt Group
Deepen Your Knowledge
Reviewed and updated by the NHIMG editorial team on June 7, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org