Microsoft Application Policy Constraints can help restrict certificate usage when the CSR is processed through Microsoft CA policy controls. EKU restrictions, by contrast, define which purposes the resulting Sub CA certificate may support and can overwrite requested usages in the CSR. For proxy and SSL intercept use cases, EKU restrictions are the more direct way to narrow certificate authority behavior.
How Microsoft Application Policy Constraints differ from EKU restrictions
Microsoft Application Policy Constraints and EKU restrictions both shape how a certificate can be used, but they act at different layers of the issuing process. Application Policy Constraints are policy controls applied by Microsoft CA during CSR processing, so they can narrow what the request is allowed to become. EKU restrictions define the permitted usages on the issued Sub CA certificate itself, which makes them the more direct control when you need to limit proxy or SSL intercept authority.
The practical difference is that Application Policy Constraints are mainly about translating or constraining what comes in from the request, while EKU restrictions are about the certificate’s final usage profile. If the CSR asks for broader purposes than you want to allow, policy constraints can shape the outcome. If you want the resulting Sub CA to be unable to issue certificates outside a narrow purpose, EKU restrictions provide the stronger and more explicit limit.
That distinction matters because a Sub CA is not just another leaf certificate. It can become a certificate authority with delegated power, so the control you choose must match where you want that power reduced. For Microsoft PKI designs, the narrower and more deterministic the issued certificate needs to be, the more you want the restriction to live in the certificate’s intended usage rather than only in request processing. For general certificate usage policy, the CA/Browser Forum baseline requirements remain a useful reference point for understanding how certificate purpose and trust expectations are governed in publicly trusted ecosystems: CA/Browser Forum.
Why EKU restrictions are usually the better fit for proxy Sub CAs
For proxy Sub CAs, the core requirement is usually not just “accept or reject this request,” but “make sure the issued subordinate CA can only support the intended interception or proxy purpose.” EKU restrictions are a direct fit for that because they bind the allowed certificate purposes on the issued Sub CA. In other words, they are closer to the actual security boundary you are trying to enforce.
Microsoft Application Policy Constraints can still be useful when you need the CA to enforce policy at issuance time, especially if you want the request path to be filtered through Microsoft CA behavior. But they are less direct when the goal is to define the operational scope of the subordinate CA after issuance. If the proxy Sub CA is later reused, cloned, or trusted in a different context, the EKU profile is the artifact most likely to travel with it and still constrain what it can do.
That is why practitioners usually treat EKU restrictions as the stronger control for purpose limitation, and Application Policy Constraints as a supporting control for issuance-time governance. The former narrows the certificate’s authority; the latter narrows what the CA will allow through the request pipeline.
What to verify before relying on either control
Before you rely on either mechanism, verify where the restriction is actually enforced and whether the effective certificate usage matches the design intent. A request-time policy can look strong on paper while the final certificate still carries broader semantics than expected. Likewise, an EKU restriction can be weakened if the resulting certificate is not validated the way you assume in the target proxy or interception workflow.
You should also confirm that the Sub CA is not being reused beyond the intended proxy scope. If the same subordinate authority can issue certificates for other purposes, the risk is no longer just configuration drift, it is authority creep. In that sense, the question is not only which control exists, but which control survives operational reuse, renewal, and administrative exceptions.
When certificate purpose, trust path, and interception behavior all need to line up, a policy-only approach is usually too indirect. A usage-bound certificate is easier to reason about, easier to audit, and harder to misapply later.
Risk and Threat Considerations
The main risk is overbroad subordinate CA authority. If a proxy Sub CA is issued with weaker purpose constraints than intended, it can become capable of signing certificates outside the interception use case, which expands the blast radius of compromise or misuse. That matters because certificate authority power is inherently high leverage, especially when the certificate is deployed in environments that implicitly trust the issuing chain.
Failure mechanism: Application Policy Constraints only shape the request and policy path, so they can fail to prevent broad effective usage if the issued certificate or downstream trust handling does not enforce the intended narrow purpose.
Impact: The Sub CA may be able to support unintended certificate issuance or interception behavior, creating broader trust abuse, harder revocation decisions, and a larger recovery problem if the CA is misused or compromised.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
NIST SP 800-53 Rev 5 and CIS Controls v8 set the technical controls, while ISO/IEC 27001:2022 defines the regulatory obligations.
| Framework | Control / Reference | Relevance |
|---|---|---|
| NIST SP 800-53 Rev 5 | SC-17 — Public Key Infrastructure Certificates | Covers certificate issuance and constrained certificate use in PKI. |
| Recommendation — Restrict subordinate CA certificate purposes to the minimum set needed for the proxy role. | ||
| ISO/IEC 27001:2022 | A.8.24 — Use of cryptography | Applies because the question is about certificate usage constraints and trust scope. |
| Recommendation — Define and enforce certificate purpose constraints in your cryptographic usage rules. | ||
| CIS Controls v8 | CIS-5 — Account Management | Supports governance over certificate authority accounts and delegated authority. |
| Recommendation — Limit who can issue or modify subordinate CA templates and purpose restrictions. | ||
Practitioner Guidance
What to prioritise: Treat EKU restrictions as the primary control when the business requirement is to constrain what a proxy Sub CA can actually do after issuance. Use Microsoft Application Policy Constraints as an issuance-time safeguard, not as the main control boundary, when the goal is strict usage narrowing.
What to verify: Check the final certificate template, the effective EKU set, and the trust path as it will be consumed by the proxy or interception platform. If the certificate can still be interpreted more broadly than intended, the design is not tight enough.
Practitioner takeaway: For proxy Sub CAs, the most important judgement is whether you are controlling request approval or controlling certificate authority behavior itself, because those are not the same security outcome.
Related resources from NHI Mgmt Group
- What is the difference between relying on application-native authentication and using a network-based identity proxy for access control?
- What is the difference between privilege reduction and secret rotation?
- What is the difference between a rules-based secret scanner and a hybrid scanner?
- What is the difference between code scanning and runtime identity monitoring?