Should organisations prioritise legacy systems when fixing orphaned accounts?

Why legacy environments deserve first pass attention

Legacy systems tend to be the places where deprovisioning breaks down: account ownership is unclear, automation is incomplete, and revocation often depends on manual steps that were never made resilient. Prioritising them is not about giving old platforms special treatment, it is about removing the most likely hiding places for stale access before you assume the central directory tells the full story.

That matters especially in acquired applications, mainframe-adjacent estates, and older infrastructure that may still trust local accounts, cached entitlements, or one-off administrative patterns. A clean identity platform can mask the fact that the legacy layer still honours access the business forgot it had.

Legacy estates also create a sequencing problem. If you start with modern systems, you may spend time proving controls that are already strong while the highest-risk orphaned access remains untouched in the oldest environment. The correct order is usually to find where access is hardest to see, then work outward.

What makes orphaned accounts harder to fix in older platforms

Older systems often lack reliable joiner-mover-leaver integration, consistent API coverage, or good inventory data, so orphaned accounts are harder to detect and even harder to revoke cleanly. In practice, this means the account may survive because nobody can confidently map it to an owner, not because it is intentionally permitted.

Legacy access models also tend to blur the difference between a dormant account, a shared account, and a service credential used by a business process. That ambiguity is dangerous because teams may postpone action until they can fully classify every account, and delay is exactly what allows stale access to persist.

Where legacy platforms are tied to business-critical functions, teams sometimes preserve access “just in case” during migrations or support transitions. That is understandable operationally, but it also makes revocation drift more likely unless the environment is explicitly queued for cleanup and reviewed as part of the migration plan.

How to decide what to fix first

Start with systems that combine weak observability, broad access, and low confidence in ownership. Those are the environments most likely to contain orphaned accounts that would otherwise evade normal access review cycles, and they are the ones most likely to make a bad situation hard to recover from later.

Prioritise platforms where a forgotten account could still reach production data, administrative functions, or integration paths used by other systems. If a legacy account can still authenticate to something valuable, the question is not whether it is old, but whether it is still materially capable of causing harm.

For a practical identity lifecycle view, teams can use the NHI Lifecycle Management Guide to structure discovery, rotation, offboarding, and visibility work, and the Joiner-Mover-Leaver (JML) Guide to keep revocation from depending on ad hoc follow-up after staff or systems change.

Risk and Threat Considerations

Orphaned accounts in legacy systems are attractive because they often sit outside modern monitoring, do not inherit current policy defaults, and may keep working long after the owning team believes they are gone. That creates a quiet persistence path for abuse, especially where a stale account still has elevated access or sits inside an acquired application that was never fully rationalised.

Failure mechanism: Revocation succeeds in the primary identity platform, but the legacy application, local directory, or embedded account store still accepts the old credential or entitlement. The result is a split-brain access state where governance looks complete while effective access still exists.

Impact: An attacker, former employee, contractor, or forgotten integration can continue using access that reviewers believe was removed. Over time, that can enable unauthorised access, privilege misuse, lateral movement, or simply a long-lived blind spot that weakens confidence in the entire access review process.

Practitioner Guidance

What to prioritise: Put the oldest, least well-instrumented, and most business-critical systems at the front of the queue. Those are the places where one orphaned account is most likely to remain effective even after central cleanup work has finished.

What to verify: Do not trust a successful directory review alone. Verify that the legacy platform itself no longer honours the account, that any local account store has been cleaned up, and that no fallback path, shared admin credential, or batch integration still depends on it.

Decision rule: If the account can authenticate anywhere outside the modern identity control plane, treat the legacy target as the source of truth for removal and confirmation, not the central directory.

Practitioner takeaway: The safest order is usually to clean the least visible systems first, because that is where orphaned access survives longest and where “we already removed it” is most likely to be false.

Related resources from NHI Mgmt Group

• How should teams reduce the risk of orphaned service accounts and stale tokens?
• When should organisations prioritise remediation of legacy data over continuing to maintain old systems?
• When should organisations prioritise distributed application architectures over incremental changes to legacy systems?
• How can organisations reduce the risk of stale API keys and machine tokens?