The control plane shift at re:Invent 2025 reshapes identity governance

At a glance

What this is: AWS re:Invent 2025 is presented as the moment identity became the control plane for humans, services, and AI agents, forcing a reset in how security architecture is organised.

Why it matters: This matters because IAM, PAM, NHI, and emerging agentic AI programmes now share the same governance surface, so identity teams need a common operating model instead of separate controls for each actor type.

👉 Read Ory's analysis of the control plane shift at AWS re:Invent 2025

Context

Identity governance is shifting because the same access fabric now has to govern people, service accounts, and AI agents inside one operational model. The article’s core claim is that architectures built for the last generation of software are too static for runtime decision-making, delegated access, and machine-speed execution.

For IAM and NHI teams, the practical issue is not whether identity matters, but where identity control sits in the stack. When identity becomes the control plane, entitlement design, policy enforcement, and runtime authorisation become inseparable, and that creates new pressure on lifecycle governance, zero trust design, and agent oversight.

Key questions

Q: How should security teams govern identity as a control plane?

A: Security teams should treat identity as the layer that decides who can act, how far authority travels, and what context makes an action legitimate. That means combining IAM, PAM, and NHI lifecycle controls with continuous evaluation, machine-readable policy, and delegation tracking. If identity is only provisioned and not governed, the rest of the stack is protecting decisions that were never validated.

Q: Why do traditional IAM controls struggle with autonomous AI agents?

A: Traditional IAM assumes predictable users or static machine accounts, but AI agents can act independently, interact with multiple systems, and generate new access needs over time. That makes static role assignment and one-time approval insufficient. Security teams need continuous evaluation, not just initial authentication and authorization.

Q: What breaks when machine identity and delegated authority are treated as the same thing?

A: You lose the ability to distinguish authentication from permission to act. A workload or agent may authenticate correctly but still require narrow tool access, tighter data scope, and explicit runtime constraints. If governance treats every non-human identity as interchangeable, privilege creep becomes invisible until behaviour changes outside the original trust boundary.

Q: Which frameworks help teams evaluate identity governance and zero trust together?

A: NIST Cybersecurity Framework 2.0 and the OWASP Non-Human Identity Top 10 are useful starting points because they connect governance, access control, and identity risk. Teams should use them to compare lifecycle coverage, recovery assurance, and least-privilege enforcement across human and non-human identity processes.

Technical breakdown

Identity as the control plane

A control plane is the decision layer that tells systems what they may do, when they may do it, and under what policy. In this model, identity is no longer only about authentication or login. It becomes the mechanism that binds subject, workload, and action together across human sessions, service-to-service calls, and AI agent execution. That shift matters because privilege is increasingly evaluated at runtime rather than only at provisioning time. The architecture therefore needs policy, trust, and telemetry to work as one system instead of three separate governance layers.

Practical implication: identity teams need to treat runtime authorisation as a core control-plane function, not a downstream access check.

Why legacy IAM assumptions break down

Legacy IAM assumes access can be defined, reviewed, and certified in relatively stable terms. That works poorly when the actor is a workload or agent that may select tools, call APIs, and shift context inside the same session. The article points to a broader Gen III security transition in which fixed-perimeter thinking gives way to identity-driven orchestration. For NHI governance, the challenge is that static policy snapshots do not describe dynamic behaviour well enough to control blast radius or delegation chains.

Practical implication: organisations should map which access decisions still rely on static assumptions and which now require continuous evaluation.

Agent IAM and machine-to-machine trust

Agent IAM extends traditional machine identity problems because the subject is not just a workload, but a runtime actor making decisions across tools and data sources. That raises the bar for how credentials, permissions, and audit trails are assigned. OIDC, API keys, secure API access, and federated identity patterns remain relevant, but they need tighter governance when agents can act on behalf of users or services. The key architectural point is that tool access and identity trust can no longer be separated cleanly in agentic environments.

Practical implication: security architects should align agent access with explicit runtime trust boundaries, not just with authentication success.

NHI Mgmt Group analysis

Identity governance is becoming control-plane governance, not just access governance. That is the real significance of this article’s thesis. When humans, services, and AI agents share one operational plane, identity policy starts shaping system behaviour in real time rather than merely recording who signed in. IAM leaders should read this as a programme design shift, not a tooling trend.

The old assumption that privileges can be fixed at provisioning time is breaking down. Provisioning-time least privilege was designed for actors whose intent was known before execution began. That assumption fails when AI agents and runtime services can change tool use, scope, and sequencing during operation. The implication is that entitlement design has to account for behaviour, not only role assignment.

Agent identity collapses the boundary between workload access and delegated decision-making. A service account used for a deterministic workload is one thing. An agent that can choose tools, retrieve context, and trigger actions is something else entirely. That is why the governance model must distinguish machine authentication from machine authority. Practitioners should not treat all non-human identities as equivalent.

Runtime authorisation drift is the right concept for this transition. The article points to a world where identity is continuously shaping action across multiple actor types, so authority can drift away from the original trust intent even when authentication is sound. That affects NHI governance, PAM design, and agent oversight in the same control surface. Security teams need a model that tracks lived behaviour, not just issued entitlements.

Cross-domain identity governance is now the differentiator. The most useful programmes will be the ones that unify human IAM, NHI controls, and agentic policy into one lifecycle view. Fragmented ownership across IAM, cloud, and AI teams will miss the handoffs where authority expands. The practical conclusion is that governance must follow the actor across the full delegation chain.

From our research:

• The average estimated time to remediate a leaked secret is 27 days, despite 75% of organisations expressing strong confidence in their secrets management capabilities, according to The State of Secrets in AppSec.
• Only 44% of developers are reported to follow security best practices for secrets management, showing that control design and control behaviour are still misaligned.
• For a broader lifecycle view, Ultimate Guide to NHIs , 2025 Outlook and Predictions helps position secret governance inside a wider identity programme.

What this signals

Runtime authority is becoming the real unit of risk. As identity shifts into the control plane, security teams should expect older review cycles to miss short-lived but high-impact access paths. That makes continuous visibility across humans, NHIs, and AI agents more important than periodic attestation alone.

With 43% of security professionals already concerned about AI systems learning and reproducing sensitive information patterns from codebases, the governance conversation is moving from access only to behaviour plus memory. Teams that own identity, secrets, and AI oversight separately will struggle to see the full pattern.

The strongest programmes will be the ones that connect policy, telemetry, and lifecycle in one operating model. For teams building that path, the NIST SP 800-207 view of continuous verification is a useful anchor point for runtime governance.

For practitioners

• Map identity control-plane responsibilities Document which team owns authentication, authorisation, lifecycle review, and runtime policy for humans, service accounts, and agents. The goal is to eliminate gaps where no group owns the decision layer.
• Separate machine authentication from machine authority Review non-human identities to confirm that successful authentication does not automatically imply broad tool use, data access, or action execution. Apply stricter policy boundaries where actors can make runtime choices.
• Inventory delegated access paths end to end Trace how access flows from users to services to automated systems and agentic components, including any handoffs across clouds, APIs, and MCP-connected tools. Focus on where authority can expand without a new approval step.
• Rework lifecycle reviews for dynamic actors Adjust access review and recertification processes so they evaluate behaviour, delegation, and runtime scope, not just static entitlements. This is especially important for AI agents and high-risk service accounts.

Key takeaways

• Identity is no longer just an access layer when humans, services, and AI agents all depend on the same control plane.
• Static provisioning assumptions are weakening because runtime behaviour now determines how far authority actually travels.
• IAM teams need one governance model for human identity, NHI, and agentic access, or delegation risk will outpace review cycles.

Key terms

• Control Plane: The control plane is the set of actions that create, configure, or manage a service. For AI workloads, it covers deployment and administration of the model platform, while data-plane permissions govern what the service and its identities can read or process.
• Runtime Authorisation: Runtime authorisation is the practice of deciding access while a task is in progress, rather than only at provisioning time. It matters for NHIs because credentials and entitlements can change risk mid-session, especially when automation or AI agents interact with sensitive systems.
• Delegation Chain: A delegation chain is the sequence of identities, credentials, and tool calls an agent uses to complete a task across systems. It matters because each step may appear acceptable on its own while the combined path produces an outcome no reviewer would have approved directly.
• Access-Centric IAM: Access-centric IAM treats access as a lifecycle process rather than a static entitlement. It links issuance, renewal, usage, and removal so security teams can govern human and non-human identities with the same operating logic across hybrid environments.

What's in the full article

Ory's full blog post covers the operational detail this post intentionally leaves for the source:

• The article’s framing of Gen III security and how the vendor connects it to identity becoming the control plane.
• The specific platform and ecosystem references around Ory Agent Security, Ory MCP Server, and machine-to-machine auth.
• The surrounding product and ecosystem context that explains how the vendor is positioning agentic AI and IAM together.
• The exact event and source material from AWS re:Invent 2025 that informed the post’s viewpoint.

👉 The full Ory post covers the ecosystem framing, platform references, and re:Invent context behind the control-plane shift.

Deepen your knowledge

NHI governance, agentic AI identity, and machine identity security 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.