TL;DR: P0 Security argues that cloud-native environments have outgrown vault- and bastion-led PAM because modern access is mediated by APIs, entitlements, ephemeral tokens, and thousands of non-human identities, making short-lived, least-privileged, auditable access the practical baseline. Legacy privileged access assumptions break when identity and access paths are dynamic rather than static.
At a glance
What this is: This is an argument for API-led privileged access management in cloud-native environments, with the key finding that traditional vault- and bastion-based PAM no longer fits how modern access is issued and revoked.
Why it matters: It matters because IAM, PAM, and NHI teams now have to govern ephemeral, API-mediated privilege across humans, service accounts, and AI agents without relying on static credential models.
👉 Read P0 Security's analysis of API-led PAM for cloud-native privilege
Context
Cloud-native privilege is no longer centred on a small number of static administrator accounts. The article argues that modern environments now rely on APIs, entitlements, ephemeral tokens, service accounts, NHIs, and increasingly AI agents, which means the old PAM model has to be judged against how access is actually created and revoked today.
The governance gap is not just technical sprawl. Traditional PAM was built around passwords, SSH keys, vaults, and bastions, while cloud-native infrastructure exposes privilege through native service APIs and short-lived permissions that do not behave like classic human admin access. That changes how least privilege, auditability, and offboarding need to work.
Key questions
Q: How should security teams implement just-in-time privileged access in cloud environments?
A: Start with the most sensitive administrative paths, then require approval, session bounds, and automatic expiry for each elevation event. JIT works best when it is paired with central policy, strong logging, and a clear offboarding path for temporary rights. The goal is to make privilege a short-lived state, not a persistent account property.
Q: Why do vaults and bastions fail to control modern privileged access?
A: Vaults and bastions were designed for static secrets and mediated sessions, but cloud-native privilege is often defined by entitlements, tokens, and API-authorized actions. That means the control can protect the login path while leaving standing permissions untouched, which is where real abuse happens.
Q: What breaks when privileged access is still standing instead of task-scoped?
A: Standing privilege gives attackers immediate value the moment a credential is stolen. They can move from access to control without waiting for approval, because the account already carries elevated rights between sessions. That is why modern PAM shifts privileged access to just-in-time, session-scoped credentials that expire automatically after the task ends.
Q: Why do service accounts and AI agents need different controls from human users?
A: Service accounts and AI agents authenticate and act without the predictable patterns that human identity systems expect. They can operate across runtimes, scale quickly, and carry permissions into automated workflows. That means access decisions should consider workload context, runtime behaviour, and time-bound authority rather than relying only on user-centric IAM patterns.
Technical breakdown
Why vault-led PAM breaks in cloud-native environments
Vault-led PAM assumes privileged access is mediated by static credentials that can be stored, rotated, and reissued. In cloud-native systems, privilege is often expressed as entitlements, tokens, role assumptions, and API-authorized actions rather than a password sitting behind a vault. That makes the vault a partial control at best, because it protects the secret but not the full lifecycle of the permission itself. When access is granted through cloud-native control planes, the real governance object is the entitlement and its issuance path, not just the credential material.
Practical implication: treat entitlement lifecycle as the control surface, not only secret storage.
Why bastion-led access does not equal just-in-time privilege
Bastion-led models insert an intermediary jump path, but that does not automatically eliminate standing privilege. If the underlying session is long-lived, key-based, or reused across tasks, the architecture still carries persistent access assumptions. In practice, this leaves a gap between the visible session broker and the actual authority granted in downstream systems. JIT access only exists when the privilege is provisioned for the task, constrained to the minimum scope, and revoked when the work ends, regardless of how the operator reached the environment.
Practical implication: verify whether the control removes privilege or merely routes it through a proxy.
How API-led PAM maps privilege to native cloud control planes
An API-led PAM model uses the platform's own APIs to create, scope, and revoke access across cloud databases, Kubernetes, virtual machines, and other infrastructure services. That matters because modern environments expose security controls programmatically, so privilege can be managed at the same layer where workloads operate. The architecture also supports short-lived access by design, which reduces the window in which stolen or over-assigned privileges can be abused. In governance terms, this is a move from credential-centric administration to machine-readable authorization and lifecycle control.
Practical implication: integrate PAM with native cloud APIs so access can be issued and removed at the same authority layer.
Threat narrative
Attacker objective: The attacker aims to turn one compromised privilege path into broad access across cloud-native infrastructure and the sensitive systems it governs.
- Entry begins when attackers obtain exposed static credentials, overprivileged entitlements, or reusable access tokens from cloud-native environments.
- Escalation follows when those credentials grant broader permissions than the task required, allowing movement from one resource type to another.
- Impact occurs when standing access or weak offboarding lets the attacker reach databases, clusters, or administrative control planes without needing to break additional authentication steps.
Breaches seen in the wild
- Azure Key Vault Contributor escalation 2024: Datadog found Azure Key Vault Contributor could add itself to access policies and read every secret, key and certificate in a vault.
- BeyondTrust breach 2024: A stolen BeyondTrust Remote Support API key let a China state-sponsored actor reset accounts and reach US Treasury workstations in 2024.
Read and download The State of NHI & AI Agent Breach Report 2026, covering 200+ breaches impacting Non-Human Identities including AI Agents.
NHI Mgmt Group analysis
API-led PAM is a response to privilege becoming a programmable control problem. Cloud-native environments no longer expose access mainly through a few human-managed credentials; they expose it through APIs, entitlements, tokens, and workload-linked permissions. That shift means the governance question is no longer only who has a password, but which systems can mint, scope, and revoke privilege at runtime. Practitioners should treat privilege orchestration as part of identity architecture, not an add-on admin workflow.
Vault-led and bastion-led models both preserve older assumptions about how privilege behaves. They were designed around static secrets or mediated sessions, which made sense when access paths were narrower and more predictable. In cloud-native estates, however, entitlement state is the real control object, and standing access can hide behind apparently controlled entry points. The implication is that PAM programmes need to be evaluated against entitlement lifecycle, not against whether a secret is stored in one place.
Short-lived access is now a governance requirement, not just an efficiency preference. The article's core tension is that development velocity and secure administration have been treated as opposing goals. In practice, the better model is to remove the trade-off by making privilege task-scoped, automatically revoked, and auditable through the same native control plane that granted it. Practitioners should view JIT and least privilege as operational defaults for cloud-native access rather than exceptional workflows.
Privileged access is increasingly shared by humans, service accounts, and AI agents, which makes lifecycle governance cross-actor by default. When the access path is API-mediated, the same entitlement mechanics can govern administrators, non-human identities, and autonomous systems. That means access review, offboarding, and audit evidence all need to speak the language of permissions and issuance events, not just user records. Teams should redesign PAM so the governance model can follow the identity type, not the other way around.
Ephemeral entitlement control is the named concept that best captures this shift. The old problem was protecting secrets; the new problem is governing access that exists only long enough to complete a task and then disappears. That changes both the attack surface and the audit model because the control has to operate at issuance time, not after the fact. Practitioners should measure PAM maturity by how well they can create and remove task-scoped privilege across native APIs.
From our research library:
- 97% of NHIs carry excessive privileges, increasing unauthorised access and broadening the attack surface, according to the Ultimate Guide to NHIs.
- Read next: Cloud PAM and CIEM Guide
What this signals
Ephemeral entitlement control: cloud-native PAM programmes need to shift their primary unit of governance from the credential to the permission event. That change matters because a vault can store a secret without proving that the secret should still confer access; the control has to live where the entitlement is created and destroyed.
Modern PAM programmes should be measured by whether they can issue least privilege through native APIs without introducing standing access through a proxy layer. If the access path still depends on long-lived credentials or reusable session grants, the architecture is preserving the old risk model under a new interface.
For practitioners
- Define privileged access around entitlements, not just credentials Map where cloud databases, Kubernetes clusters, VMs, and service accounts receive authority through API permissions rather than stored passwords or SSH keys.
- Replace standing access with task-scoped issuance Require access to be provisioned for a specific job and automatically removed when the task ends, even when the request came through a proxy or jump path.
- Audit the lifecycle of privileged permissions Review who can create, extend, and revoke high-risk entitlements in native control planes, and document where the authority sits outside traditional vault tooling.
- Unify PAM governance across humans and NHIs Use the same approval, expiry, and offboarding logic for administrators, service accounts, and AI-driven access flows so governance matches the actual identity type.
Key takeaways
- Cloud-native privilege now flows through APIs, entitlements, tokens, and workload identities, which makes classic secret-centric PAM incomplete.
- The article's central governance point is that task-scoped access has to be created and revoked in the native control plane, not merely brokered through a vault or bastion.
- PAM teams should treat humans, service accounts, and AI-driven access flows as part of one entitlement lifecycle instead of separate administration problems.
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 addresses the attack and risk surface, while NIST SP 800-53 Rev 5, NIST CSF 2.0, NIST Zero Trust (SP 800-207) and CIS Controls v8 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 centers on excessive privilege across NHIs, service accounts, and cloud entitlements. |
| NHI-07 — Long-Lived Secrets | It contrasts static credentials with ephemeral, task-scoped access in cloud-native estates. | |
| NHI-10 — Human Use of NHI | The post warns that human administration patterns do not map cleanly onto API-driven non-human access. | |
| Recommendation — Reduce standing authority for NHIs and align access issuance to least privilege at the entitlement layer. Replace long-lived privileged secrets with short-lived credentials and automated revocation. Separate human administration from machine-issued privilege and govern each through its own lifecycle. | ||
| NIST SP 800-53 Rev 5 | IA-5 — Authenticator Management | IA-5 governs credential lifecycle controls, including issuance and revocation. |
| Recommendation — Apply authenticator lifecycle controls so privileged credentials expire and are revoked on task completion. | ||
| NIST CSF 2.0 | PR.AA-05 — Access Permissions, Entitlements and Authorizations | The article is fundamentally about entitlement-based access in cloud-native systems. |
| Recommendation — Govern access by permissions and authorizations, not by static credential possession alone. | ||
| NIST Zero Trust (SP 800-207) | Zero Trust architecture principles — Zero Trust architecture principles | Just-in-time privilege and continuous verification are core zero-trust access patterns. |
| Recommendation — Use zero trust principles to remove standing access and require verified, task-scoped authorization. | ||
| CIS Controls v8 | CIS-5 — Account Management | The article focuses on lifecycle control for privileged accounts and non-human identities. |
| Recommendation — Centralise account governance so privileged access can be provisioned, reviewed, and revoked consistently. | ||
Key terms
- API-Led PAM: API-led PAM is a privileged access model that enforces request, approval, granting, and revocation through application and infrastructure APIs rather than through separate intermediary tooling. It is designed for dynamic production environments where access must be automated, scoped, and auditable across the full lifecycle.
- Session-scoped privilege: Access that exists only for the current task or execution window and is removed when the session ends. For autonomous or agentic systems, this reduces standing privilege but also shifts the burden to runtime controls, because the identity may not persist long enough for traditional review cycles.
- Standing Access: Standing access is persistent privilege that remains available without fresh approval or contextual checks. In NHI environments, standing access usually appears as long-lived tokens, reusable service accounts, or broad roles attached to automation. It is convenient operationally, but it expands risk when conditions change or secrets leak.
- Entitlement Lifecycle: The entitlement lifecycle covers how access is created, reviewed, used, changed, and removed over time. Strong lifecycle control prevents old permissions from lingering after a role, project, or need has ended, which is essential for least privilege and audit readiness.
What's in the full article
P0 Security's full article covers the operational detail this post intentionally leaves for the source:
- How the API-led PAM model is applied across specific cloud services and infrastructure types
- The workflow mechanics for provisioning and revoking privileged access through native APIs
- The article's practical discussion of why static credentials and SSH-based access no longer fit cloud-native control paths
- The compliance angle behind identity inventories and access controls in SOC 2, FedRAMP, and ISO 27001
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 responsible for identity security strategy or NHI governance in your organisation, it is worth exploring.
Published by the NHIMG editorial team on October 13, 2026.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org