Infrastructure identity for agentic IT: why control plane thinking matters

At a glance

What this is: This is a Teleport analysis of infrastructure identity, arguing that cryptographic, short-lived access should anchor human, machine, and AI-driven actions across modern enterprise infrastructure.

Why it matters: It matters because IAM teams now have to govern not just people, but workloads and AI agents that act at machine speed, where standing privilege and reusable credentials create avoidable exposure.

By the numbers:

• Only 5.7% of organisations have full visibility into their service accounts.
• NHIs outnumber human identities by 25x to 50x in modern enterprises.
• 91.6% of secrets remain valid five days after the targeted organisation is notified, showing a critical gap in remediation procedures.
• When AWS credentials are exposed publicly, attackers attempt access within an average of 17 minutes , and as quickly as 9 minutes in some cases.

👉 Read Teleport's analysis of infrastructure identity for agentic IT

Context

Infrastructure identity is the idea that every actor reaching critical systems should carry a cryptographic identity into the session, rather than relying on reusable secrets or anonymous network trust. Teleport's article argues that this model becomes necessary as workforce identity, service accounts, automation, and AI agents all converge on the same operational surface.

The governance gap is that traditional PAM and vault-based controls were built for slower, human-paced administration. Once operators or agents route around those controls to keep work moving, organisations lose the ability to distinguish who acted, with what privilege, and for how long.

For identity teams, the key question is no longer whether access exists, but whether the access model can bind actor context to each action in real time. That is where NHI governance, workload identity, and agent-ready access controls now intersect.

Key questions

Q: How should security teams govern AI agents that can access enterprise systems?

A: Security teams should govern AI agents as non-human identities with explicit ownership, scoped privileges, and continuous monitoring. The control set should include inventory, task-bound credentials, audit trails, and revocation paths. If an agent can call tools or touch production systems, it belongs in the same governance model as service accounts and other machine identities.

Q: Why do standing admin credentials create more risk in modern environments?

A: Standing admin credentials create more risk because they extend the time window in which an account can be abused, misused, or forgotten. In cloud-heavy estates, that risk compounds when credentials are shared across tools or left behind after a role change. Short-lived access reduces exposure, but only if revocation and audit coverage are reliable.

Q: What breaks when organisations rely on password vaults for every privileged identity?

A: Password vaults still help, but they break down when the real risk is persistent authorisation rather than secret storage. If an identity can access a system through certificates, tokens, APIs, or agentic workflows, the vault does not fully govern what that identity can do at execution time. That leaves the privileged action itself under-controlled.

Q: Who is accountable when a compromised AI agent misuses delegated access?

A: Accountability usually spans the business owner of the workflow, the team that issued or approved the credential, and the vendor if a third-party integration was involved. The critical governance question is not who logged in, but who allowed the delegation chain to exist and remain valid. That chain must be documented before incidents occur.

How it works in practice

Why credential custody breaks under infrastructure identity

Credential custody treats secrets as the control point: a vault stores them, rotates them, and releases them when a workflow needs access. That approach works poorly when the real problem is actor accountability at the moment of action. Infrastructure identity instead binds an actor's cryptographic identity to the session and issues task-scoped privilege that expires with the work. The architecture shifts from secret distribution to verifiable session issuance, which is much harder to bypass without creating an auditable trace. In practice, the weak point is not encryption, but the assumption that holding a secret is the same as governing an actor.

Practical implication: replace secret-centric access paths with session-bound identity and short-lived privilege for infrastructure operators and automation.

How standing privilege expands the NHI blast radius

Standing privilege is persistent access that remains valid beyond any single task or maintenance window. In infrastructure environments, that creates a large blast radius because one compromised credential can be reused repeatedly, copied into tooling, or routed through alternate paths when the approved process is slow. The article's model pushes toward access that is issued only for the task at hand and ends automatically when the session closes. This does not merely reduce exposure time. It also changes what investigators can reconstruct because every privileged action is tied to a discrete identity and session record.

Practical implication: identify infrastructure accounts that retain access outside maintenance windows and reclassify them for just-in-time issuance.

Why AI agents need the same actor-bound governance

AI agents complicate infrastructure access because they can act continuously and at machine speed, but they still need identity, privilege, and session boundaries. Teleport's framing is that agents should not inherit broad access simply because they are efficient or because a human asked them to operate. If the governance model cannot attribute actions to a specific actor and window of authority, then agent activity becomes indistinguishable from ambient automation. The technical issue is not whether the agent is intelligent. It is whether the system can enforce a narrow, attributable path for each action the agent performs.

Practical implication: put AI agents into the same identity and session governance path as workloads, with explicit task-scoped access and attribution.

NHI Mgmt Group analysis

Infrastructure identity is the next governance layer because identity now sits inside the control plane. The article is right to treat network location as insufficient once human operators, workloads, and AI agents all reach the same infrastructure surface. IAM programmes that still rely on perimeter context are already behind the operating model they need to govern.

Standing privilege is the real failure mode, not just weak secrets management. Traditional vaulting reduces exposure, but it does not solve the governance problem when access remains valid after the task is complete. This is the same structural weakness that appears across NHI programmes: access outlives intent, so accountability becomes ambiguous before remediation even begins.

Infrastructure identity changes the least-privilege question from provisioning-time design to runtime actor binding. Least privilege was designed for actors whose purpose is known before execution begins. That assumption weakens when automation and AI agents can change what they do mid-session, so the implication is that entitlement review alone no longer describes the control problem.

Agent-ready infrastructure governance requires the same discipline used for NHI, but with tighter session semantics. Workloads and AI agents are both non-human actors, yet agents can introduce rapid path changes that make static access models even less reliable. Practitioners should treat this as a convergence point for NHI governance, PAM, and emerging agent control models, not as a separate niche.

Actor context, not IP address, is becoming the meaningful policy input. The article points toward identity-aware infrastructure where connection metadata alone is not enough to determine trust. That aligns with Zero Trust thinking, but it also raises the bar for NHI lifecycle management because every actor now needs revocation, attribution, and expiry that match its operational window.

From our research:

• Only 5.7% of organisations have full visibility into their service accounts, according to Ultimate Guide to NHIs.
• 91.6% of secrets remain valid five days after the targeted organisation is notified, according to Ultimate Guide to NHIs, which shows why expiry lag is a control weakness.
• 52 NHI Breaches Analysis shows how lifecycle gaps and missing revocation turn access into long-lived exposure.

What this signals

Infrastructure identity will force identity teams to treat the access path itself as governed infrastructure. That shift matters because the old separation between network control and identity control is collapsing in AI-heavy environments. Practitioners should expect tighter coupling between PAM, workload identity, and session telemetry as the simplest way to preserve attribution and reduce alternate-path risk.

Identity teams should expect NHI lifecycle management to absorb more infrastructure administration than classic PAM ever covered. Once cryptographic access replaces shared secrets, offboarding, review, and expiry become continuous operational tasks rather than periodic clean-up. Organisations that cannot inventory non-human actors will struggle to enforce actor-bound policy at the pace modern infrastructure now demands.

Short-lived privilege is becoming the practical baseline for agentic IT. As AI agents and automation take on more work, access windows will need to shrink until they match the life of the task, not the convenience of the operator. That is where identity-aware infrastructure starts to matter operationally, not just architecturally.

For practitioners

• Bind infrastructure access to cryptographic session identity Replace reusable secrets with identity-bound sessions for administrators, automation, and service workflows. Require each privileged action to inherit the actor context that initiated it, and make session records the authoritative audit trail.
• Eliminate standing privilege outside maintenance windows Inventory infrastructure accounts that retain access after the task ends, then move them to just-in-time issuance with automatic expiry. Prioritise accounts that can reach production systems, orchestration planes, or network infrastructure.
• Bring AI agents into the same access model as workloads Do not let agents inherit broad administrative access because they are fast or convenient. Assign task-scoped permissions, constrain tool reach, and ensure every action is attributable to a specific agent session.
• Reduce alternate access routes that bypass governance Close the workarounds operators use when the approved path is too slow, because those paths often reintroduce standing access and hidden secrets. Measure how often teams leave the vault to get work done.
• Test blast-radius assumptions against machine-speed actors Use incident reconstruction exercises to validate whether you can trace privileged actions back to the actor, the session, and the privilege window. If you cannot, your governance model is still too permissive for agentic IT.

Key takeaways

• Teleport's core argument is that infrastructure identity must replace secret-centric access models because actor context now matters more than network location.
• The governance failure is standing privilege, which keeps access alive after intent has changed and makes infrastructure harder to attribute and contain.
• Identity teams should move administrative, workload, and agent access toward session-bound, task-scoped privilege if they want control that still works at machine speed.

Key terms

• Infrastructure Identity: Infrastructure identity is the set of credentials and trust relationships that allow systems, workloads, and automation to authenticate to other systems. It is the machine-facing layer of identity governance, and it often carries more operational risk than human access because it is persistent and widely reused.
• 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.
• Session-bound Access: Session-bound access is access that exists only for a specific activity window and then disappears. In OT governance, it is the right pattern for remote support because it reduces persistent exposure, improves accountability, and fits the operational reality of maintenance and incident response.
• Actor Context: The identity information that explains who or what is acting, under what authority, and for which purpose. In infrastructure governance, actor context matters because connection metadata alone cannot reliably distinguish legitimate administration from unsafe or inherited access.

What's in the full announcement

Teleport's full analysis covers the operational detail this post intentionally leaves for the source:

• How Teleport expects short-lived cryptographic access to replace vault-mediated workflows in infrastructure administration.
• The specific operator journey for entering a maintenance window without retrieving a reusable credential.
• How Cisco and Teleport describe session brokering, recording, and identity-aware enforcement at the network layer.
• What customers should expect as infrastructure teams move from password-based access to actor-bound access records.

👉 Teleport's full post covers the transition from credential custody to identity-native access in more operational detail.

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.