By NHI Mgmt Group Editorial TeamPublished 2023-12-19Domain: Best PracticesSource: EmpowerID

TL;DR: SCIM can standardise cloud provisioning, but many enterprise applications still lack native support, forcing organisations to bridge Azure to legacy and custom systems through an intermediary layer, according to EmpowerID. The governance issue is not automation itself but where policy, approval, and lifecycle control sit when identity changes must flow across non-SCIM applications.


At a glance

What this is: This is an analysis of how SCIM mediation between Azure and non-SCIM applications changes identity provisioning and governance.

Why it matters: It matters because IAM teams still have to govern provisioning, approval, and lifecycle control across mixed application estates, even when the target systems cannot speak SCIM natively.

👉 Read EmpowerID's article on SCIM mediation for Azure provisioning


Context

SCIM, or System for Cross-domain Identity Management, is meant to standardise how identity data moves between directories and applications. The problem is not the protocol design. The problem is that many enterprise systems still do not support it, which leaves IAM teams with fragmented provisioning paths and inconsistent lifecycle control across Azure and non-SCIM applications.

That creates a governance gap for NHI and human identity programmes alike, because the access change may be initiated centrally but executed through application-specific logic. When provisioning spans custom applications, ERP systems, and HR platforms, the real question becomes where policy decisions are enforced and where exceptions are approved. For teams building lifecycle controls, the relevant reference point is the NHI Lifecycle Management Guide.


Key questions

Q: How should teams govern provisioning when some applications do not support SCIM?

A: Teams should treat SCIM as the preferred standard but plan for a mediation layer where native support is missing. The key is to keep policy, approval, and audit control in one place rather than letting each downstream system handle identity changes differently. That approach reduces manual exceptions and makes provisioning behaviour easier to govern across mixed application estates.

Q: Why do non-SCIM applications create identity governance risk?

A: Non-SCIM applications create risk because they force organisations into custom integration patterns that are harder to standardise, review, and decommission. Every bespoke connector adds another place where provisioning, group membership, or offboarding can drift from policy. The issue is not the application type alone, but the loss of a consistent lifecycle control path.

Q: What breaks when provisioning is treated as a fire-and-forget integration?

A: Governance breaks first. If commands simply pass through without intermediate policy evaluation, the organisation loses a checkpoint for approval, exception handling, and naming logic. That makes it harder to explain why an access change happened, who approved it, and whether the downstream system received the right entitlement.

Q: How can IAM teams decide whether a virtual directory should be privileged?

A: If the intermediary can route identity changes to many applications, it should be handled as a privileged administrative path. IAM teams should review who can edit routing, which applications are registered, and whether service authentication is scoped tightly enough. The decision hinges on blast radius, not convenience.


Technical breakdown

SCIM provisioning as a protocol layer

SCIM is an open provisioning standard that lets one system request identity changes in another system using a common schema and REST-style calls. In practice, SCIM reduces integration variance only when both sides support the protocol and agree on object models such as users and groups. Where support is missing, organisations either build custom connectors or place a mediation layer in front of the target systems. That mediator becomes the control point for request translation, policy checks, and lifecycle orchestration.

Practical implication: treat SCIM coverage as an architecture map, not a checkbox, before assuming provisioning consistency.

Workflow-driven virtual directories and policy enforcement

A workflow-driven virtual directory does more than forward provisioning commands. It can evaluate business rules before a change is committed, which means approval, naming, password policy, and routing logic can be applied in one place rather than scattered across downstream applications. This shifts the control boundary from the application to the mediation layer. The security value comes from making provisioning decisions observable and interruptible before identity state changes propagate into connected systems.

Practical implication: place approval and policy evaluation at the mediation layer where downstream applications cannot enforce them uniformly.

Azure managed identity and intermediary trust

When a provisioning intermediary is deployed as an Azure App Service, the trust model depends on how that service authenticates to Azure and to each connected application. An Azure managed identity can reduce secret handling for the intermediary itself, but it does not remove the need to govern the applications behind it. If the mediator can update many systems, it becomes a high-value administrative path that must be tightly scoped and monitored.

Practical implication: scope the intermediary's privileges separately from the business applications it provisions.


NHI Mgmt Group analysis

SCIM adoption has become a governance problem, not just an integration problem. When applications do not support SCIM natively, identity teams lose a uniform lifecycle path and fall back to bespoke connectors, manual exceptions, or vendor-dependent timelines. That fragmentation makes provisioning consistency harder to prove and harder to audit. The practitioner conclusion is that SCIM coverage should be treated as part of identity control design, not as an optional engineering convenience.

Workflow mediation is the control boundary that matters. The article shows that the real question is not whether commands can move from Azure to a target application, but whether business logic can be enforced before those commands take effect. That is the same governance pattern IAM teams use when they separate request initiation from fulfilment. The practitioner conclusion is to define where policy is evaluated, where approvals happen, and where exceptions are allowed.

Intermediary trust concentration creates a new privilege concentration point. If one SCIM virtual directory can provision many applications, it becomes a powerful administrative bridge across the estate. That does not make it inherently risky, but it does mean privilege scope, service authentication, and change logging deserve the same scrutiny as any other high-trust identity path. The practitioner conclusion is to govern the mediator as a privileged system in its own right.

Identity lifecycle governance must span both supported and unsupported applications. Organisations often assume that provisioning standards will eventually normalise access control across the application estate, yet the long tail of legacy and custom systems remains. That means lifecycle governance has to work across heterogeneous targets without collapsing into inconsistent manual practice. The practitioner conclusion is to measure whether offboarding, group updates, and access exceptions are still traceable end to end.

From our research:

What this signals

SCIM mediation will become a governance pattern wherever application estates stay heterogeneous. Teams should expect more identity control to move into intermediary services that translate policy into fulfilment across legacy and cloud systems. That increases the importance of reviewing routing logic, approval checkpoints, and offboarding traceability in the same change window. For practitioners, the practical question is whether the mediator is auditable enough to stand in for a native control plane.

With 88.5% of organisations saying their non-human IAM practices lag behind or only match human IAM maturity, according to our 2024 Non-Human Identity Security Report, provisioning mediation is not a niche design choice. It is one of the places where those maturity gaps surface most clearly. The next step is to connect identity lifecycle governance to the actual integration path, not the intended architecture.

The strongest programmes will align SCIM strategy with the NIST Cybersecurity Framework 2.0 so that provisioning, logging, and recovery expectations are explicit. That matters because identity changes that pass through a mediation layer can obscure who approved what unless the control design is deliberate. Practitioners should watch for hidden exception paths and undocumented connectors as the estate evolves.


For practitioners

  • Map every non-SCIM dependency in the provisioning chain Document which Azure provisioning flows terminate in native SCIM targets and which rely on a mediation layer or custom connector. Use that map to identify where policy enforcement, exception handling, and offboarding logic are actually taking place.
  • Define the approval boundary before fulfilment begins Require that any workflow mediation layer evaluates policy violations, naming rules, and approval gates before it writes changes to downstream systems. Keep the fulfilment step separate from the decision step so the audit trail shows what was checked and what was allowed.
  • Treat the provisioning intermediary as a privileged system Apply tighter access scope, stronger logging, and change control to the intermediary service than to ordinary business applications. Review who can modify its routing rules, application paths, and workflow logic, because those settings determine where identity changes go.
  • Test offboarding across both native and mediated paths Verify that disabling a user, group, or workflow request removes access from all downstream applications, including systems reached through mediation. Do not assume the same lifecycle outcome just because the request originated in Azure.

Key takeaways

  • SCIM standardises provisioning only where both sides support it, so mediation becomes the real control point in mixed application estates.
  • The main governance risk is not the protocol itself but the fragmentation created when identity changes must pass through custom or non-native paths.
  • IAM teams should govern the intermediary as a privileged system and verify that approval, audit, and offboarding still work end to end.

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 address the attack and risk surface, while NIST CSF 2.0 and NIST Zero Trust (SP 800-207) set the governance and control requirements practitioners need to meet.

FrameworkControl / ReferenceRelevance
OWASP Non-Human Identity Top 10NHI-03Provisioning mediation affects lifecycle consistency across non-human identities.
NIST CSF 2.0PR.AC-4Controls access permissions across mixed application provisioning paths.
NIST Zero Trust (SP 800-207)AC-5Identity mediation should not create uncontrolled privilege concentration points.

Review whether mediated provisioning preserves the same lifecycle state changes across all targets.


Key terms

  • SCIM: System for Cross-domain Identity Management is a standard for exchanging identity data between systems in a consistent way. It helps automate provisioning and deprovisioning across applications, but only when both systems support the protocol and the organisation governs the workflow around it.
  • Virtual Directory: A virtual directory is an intermediary identity layer that translates requests between a source system and one or more target applications. In governance terms, it can become the control point where policy checks, routing, and approval logic are enforced before changes reach downstream systems.
  • Identity Provisioning: Identity provisioning is the process of creating, updating, or removing accounts and entitlements across systems. For mixed estates, the challenge is less about sending the request and more about ensuring the request follows the same approval, logging, and lifecycle rules wherever it lands.
  • Lifecycle Governance: Lifecycle governance is the set of controls that keep identity state aligned with business need from joiner through mover to leaver. It applies to humans, non-human identities, and automated identity paths, and it only works when offboarding and exception handling are traceable end to end.

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.

This post draws on content published by EmpowerID: SCIM workflow mediation for Azure provisioning and non-SCIM applications. Read the original.

NHIMG Editorial Note
Published by the NHIMG editorial team on 2023-12-19.
NHI Mgmt Group — the independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org