TL;DR: Opal Security says access governance now spans employees, service accounts, and AI agents, with Paladin, Access Campaigns, OpalScript, and MCP updates built around one access graph and a cited 43.5% of granted access going unused. The old assumption that access can be reviewed on a human cadence no longer holds when decisions and requests move at machine speed.
At a glance
What this is: This is Opal Security’s product update on GA access campaigns, policy-as-code review logic, and agent governance features built around a single access graph across employees, service accounts, and AI agents.
Why it matters: It matters because identity teams now have to govern standing access, policy drift, and machine-speed requests in the same control plane instead of treating NHI and human review as separate problems.
By the numbers:
- 43.5 percent of granted access goes unused, according to Opal Security.
👉 Read Opal Security’s product update on access campaigns and agent governance
Context
Access governance breaks down when review processes assume that access is both human-paced and stable enough to be inspected later. In this release, Opal Security is describing a control problem that now spans employees, service accounts, and AI agents within the same access graph.
The practical shift is not just more automation. It is the move from periodic review of static grants to policy enforcement over requests, campaigns, and agent-driven access paths that can change faster than manual governance can keep up.
Key questions
Q: What breaks when AI agent access is reviewed only after the fact?
A: After-the-fact review leaves a gap between action and containment. If an agent can already reach a dataset, API, or SaaS system, the damage may be done before a human sees the alert. Runtime checks reduce that gap by stopping unauthorized actions before they execute.
Q: When should teams prioritise policy-as-code over manual approval handling?
A: Teams should prioritise policy-as-code when the same access request can be judged differently by different reviewers or at different times. If consistency, auditability, and repeatability matter, code-based policy becomes the more reliable control plane because it reduces drift and makes exceptions visible.
Q: What are the signs that standing access is creating governance drift?
A: The clearest signals are unused entitlements, reviewer inconsistency, and access that persists long after the task is complete. When those patterns appear together, the programme is governing retained privilege rather than active need, and the review process is no longer shrinking blast radius.
Q: How should security teams design MCP server access for AI agents?
A: Security teams should design MCP access around a small set of agent goals, not a mirrored list of REST endpoints. Group related backend calls into workflows, expose known data as resources, and reserve tools for actions that truly change state. That keeps the agent oriented, reduces context churn, and makes the access surface easier to review and govern.
How it works in practice
Why access review workflows fail at machine speed
Access reviews were designed around a stable grant that remains visible long enough for a reviewer to assess it. When AI agents can request access, act, and move on within the same operational window, the control starts losing its reviewable state. That creates a timing mismatch between governance cadence and identity behaviour. A campaign model can reduce reviewer drag, but the deeper issue is that the decision point has shifted from certification time to issuance time.
Practical implication: move controls closer to request and issuance so the review process does not depend on later human inspection.
How policy-as-code reduces drift across access decisions
Policy drift happens when the same request gets different outcomes depending on who reviews it or when it is reviewed. Encoding access policy turns approval logic into a versioned, testable control path rather than an informal judgment layer. That matters most when access is governed across multiple identity types, because humans, service accounts, and agents all hit the same decision logic. The real value is consistency, auditability, and repeatable enforcement across changing operational conditions.
Practical implication: treat access policy as governed code so decisions remain consistent across reviewers, timestamps, and identity classes.
What MCP changes for non-human identity governance
MCP creates a standard way for agents to reach tools and data sources, which means the governance question moves from whether an agent can connect to how that connection is scoped, approved, and revoked. If the access path is approved once and left standing, the control model becomes vulnerable to privilege persistence. Scoped, expiring access is a stronger fit for agent workflows because it binds use to a defined task instead of a standing key or open-ended entitlement.
Practical implication: govern MCP access as a scoped NHI credential path with explicit expiration, not as a permanent integration credential.
NHI Mgmt Group analysis
Access governance is no longer a human-only discipline. Once agents request access through the same control plane as employees and service accounts, identity governance has to cover request timing, policy execution, and revocation across all three actor types. The discipline is shifting from certification-centric oversight to issuance-centric control, and that changes what IGA can safely assume.
Policy drift is the real control gap, not just reviewer fatigue. When the same request can receive different outcomes depending on reviewer or timing, the programme no longer has a stable enforcement model. Opal Security’s policy-as-code direction points to a larger market truth: governance is becoming executable logic, not a spreadsheet of approvals.
Standing access is now the primary blast-radius problem. The article’s 43.5% unused-access figure is a reminder that over-granted permissions create risk even before an attacker appears. In mixed human and machine environments, unused access becomes persistence capital for both misuse and compromise, so practitioners need to treat scope reduction as an operational control, not a cleanup exercise.
Named concept: access governance at machine speed. This is the point where review cycles, approval thresholds, and offboarding assumptions collide with identity behaviour that can request and consume access inside a single operational window. The implication is not just more automation; it is a redesign of governance around task-bounded, time-bounded, and actor-bounded access decisions.
One access graph is the right abstraction, but only if lifecycle rules differ by actor. Employees, service accounts, and AI agents may share the same policy plane, but they should not share the same entitlement duration, review cadence, or revocation trigger. The practitioners who succeed will separate the graph from the governance rules and make those rules explicit for each identity class.
From our research library:
- 24,008 unique secrets were exposed in MCP configuration files in 2025 alone, the protocol's first year of widespread adoption, according to the State of Secrets Sprawl 2026.
- Read next: Access Reviews and Certification Guide
What this signals
Access governance has to become actor-aware. A single control plane can still work, but only if the rules behind it distinguish between people, service accounts, and AI agents. The review cadence that is acceptable for a human user is often too slow for a non-human identity, and that gap is now a governance issue, not just an operational inconvenience.
Policy drift will matter more than feature breadth. As identity programmes expand, the main failure mode is not lack of tooling but inconsistent decisions across reviewers and channels. Organisations that can encode access rules, exceptions, and expiry logic will have a better chance of keeping blast radius under control when requests are increasingly machine-timed.
For practitioners
- Map governance rules to actor speed Review which access decisions still assume a human reviewer and a stable entitlement window. Rework those rules for agent-speed requests, where approval, execution, and revocation may need to happen within one session.
- Encode approval logic as policy Move recurring access decisions into versioned policy so the same request produces the same result regardless of reviewer or time of day. Keep exception handling explicit so reviewers can see why a rule escalated.
- Shorten the lifetime of non-human access Replace standing keys and broad agent entitlements with scoped, expiring access that matches the task. Revoke any path that does not need to survive beyond the work it was created for.
- Separate one graph from many governance rules Keep employees, service accounts, and AI agents in a shared visibility model, but assign different review cadences, duration limits, and revocation triggers to each identity class.
Key takeaways
- Access governance is moving from human-paced review to issuance-time enforcement because agents and service accounts can change state faster than manual certification cycles.
- The article’s 43.5 percent unused-access figure shows why standing privilege remains a material source of blast radius even before compromise occurs.
- The practical response is to keep one identity graph but apply different duration limits, revocation triggers, and policy logic for employees, service accounts, and AI agents.
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, OWASP Agentic AI Top 10 and MITRE ATT&CK address the attack and risk surface, while NIST CSF 2.0 and NIST SP 800-53 Rev 5 set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| OWASP Non-Human Identity Top 10 | NHI-05 — Overprivileged NHI | The article centres on unused and excessive access across non-human identities. |
| NHI-07 — Long-Lived Secrets | MCP and agent access paths need expiration instead of persistent keys or grants. | |
| NHI-10 — Human Use of NHI | The same access graph now spans employees, service accounts, and AI agents. | |
| Recommendation — Reduce standing entitlements and scope NHI access to the minimum task required. Replace persistent agent credentials with expiring access tied to task duration. Separate human and non-human governance rules even when they share one control plane. | ||
| OWASP Agentic AI Top 10 | ASI03 — Identity & Privilege Abuse | Agent-mediated access requests and approvals can amplify privilege misuse if governance is loose. |
| Recommendation — Constrain agent privileges and review every delegated access path for abuse potential. | ||
| NIST CSF 2.0 | PR.AA-05 — Access Permissions, Entitlements and Authorizations | The article is fundamentally about how access permissions are assigned and controlled. |
| Recommendation — Enforce entitlement rules that are consistent, reviewable, and time-bounded across identities. | ||
| NIST SP 800-53 Rev 5 | AC-6 — Least Privilege | Unused access and broad agent permissions are classic least-privilege failures. |
| Recommendation — Apply least privilege to every access request and remove permissions that exceed immediate need. | ||
| MITRE ATT&CK | TA0006;TA0008 — Credential Access; Lateral Movement | Standing or over-broad access expands credential abuse and movement paths. |
| Recommendation — Map excess access to credential access and lateral movement risk during control testing. | ||
Key terms
- Access Campaign: An access campaign is a structured review cycle in which designated reviewers assess a defined set of users, entitlements, or roles. Campaigns are the operational mechanism for certification at scale, usually organized around business ownership, risk priority, and deadlines that drive remediation and attestation outcomes.
- Policy as Code: Policy as code stores authorization logic in version control and evaluates it through testable, reviewable rules. For agent governance, it makes runtime decisions reproducible and measurable, which is critical when actions can be triggered by untrusted content and executed at machine speed.
- Agent Governance: Agent governance is the set of policies, controls, and evidence required to manage autonomous software as a non-human identity. It covers consent, tool access, lifecycle review, audit logging, and revocation so that an agent remains bounded as its workflows change.
- Standing Privilege: Standing privilege is access that remains active even when no immediate task requires it. For NHI programmes, it is a common failure mode because long-lived credentials and persistent roles create unnecessary exposure. Reducing standing privilege usually means tighter expiry, on-demand access, and clearer review of who or what still needs access.
What's in the full announcement
Opal Security's full product update covers the operational detail this post intentionally leaves for the source:
- How Paladin decides when to advise versus escalate during access approval
- Campaign-building workflow details for filtering, grouping, reassignment, and bulk reminders
- OpalScript examples for enforcing separation of duties and duration limits
- MCP and OAuth configuration details for scoped, expiring agent access
Deepen your knowledge
NHI governance, agentic AI identity, and machine identity lifecycle are core topics in our NHI Foundation Level course, the industry's only accredited NHI security programme. If you are building or maturing an IAM programme, it is worth exploring.
Published by the NHIMG editorial team on August 15, 2026.
Updated on October 7, 2026.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org