Automating joiner-mover-leaver processes removes delay and manual error from provisioning, but it does not reduce access risk on its own. Automation applies rules. It does not ask whether the resulting access is still appropriate. Identity governance adds that question through review, validation, and evidence, starting with the applications an audit or regulator will examine.
Most organisations that invest in Identity and Access Management (IAM) start with JML automation, and for good reason. New starters receive the access they need on their first day, people who change roles are updated without waiting in a ticket queue, and leavers lose access when HR records their exit. These are real gains in speed, consistency, and service quality. At Laurea University of Applied Sciences, automated onboarding cut student account activation from 24 hours to as little as an hour.
The difficulty arises when automation is expected to answer a different question. JML automation confirms that access was granted according to the rules. It cannot confirm that the person should still hold that access today. Access risk sits in the space between those two statements, and in most organisations that space is nobody's fault: it is not what provisioning was built to measure.
JML automation takes an event from an authoritative source, usually the HR system, and applies a set of rules to it. A joiner in the finance team receives the birthright access defined for that department. A mover receives the access defined for the new role, and a leaver has accounts disabled across the connected systems.
That makes the process only as good as three inputs: the rules, the data that triggers them, and the applications the process can reach. When all three are accurate and complete, automation performs well. When any of them drifts, automation repeats the drift consistently and at speed.
Five patterns account for most of the access risk that survives a well-run provisioning process.
Mover events are often configured to add the access a new role needs without removing what the old role granted. The reasoning is practical, because removing access on the day someone moves can interrupt handover work, so the old entitlements stay in place "for now". Without a later review, "for now" becomes permanent. After three or four moves, a long-serving employee can hold the combined access of every role they have held, which is the pattern usually described as privilege creep.
Birthright rules cover the access everyone in a role needs. They do not cover the project folder, the reporting tool licence, or the elevated permission granted to fix an urgent problem. These requests are usually approved through a ticket or directly by an application administrator, outside the logic that JML automation manages. When the person moves or leaves, the automation holds no record that the access exists, so nothing removes it.
JML automation acts only on the systems it is connected to. Many organisations connect their directory, email, and a handful of core applications, while software as a service (SaaS) tools, legacy platforms, and locally managed applications remain outside. Access in those applications depends on manual requests and manual removal, and those applications are frequently where audit findings appear.
An orphaned account has no current owner. The person it belonged to has left, or the account was never linked to an identity in the first place. Service accounts, test accounts, and accounts created directly inside an application are common examples, and because they have no matching HR record, a leaver event can never reach them.
Unused accounts carry a similar risk. The owner still works for the organisation but has not signed in for months, and the access remains available to anyone who obtains the credentials.
Some risks appear only when two entitlements are held together. The ability to create a supplier and the ability to approve payments to that supplier may each be reasonable on their own. Held by one person, they break a basic segregation of duties (SoD) principle. Provisioning rules evaluate each entitlement as it is granted, and they rarely check how it combines with everything the person already holds.
The same patterns appear in every sector, and the sector shapes which one dominates: people holding several roles at once in higher education, seasonal and supplier access in retail, and rotating clinical staff in healthcare.
The two tests sound alike, but they answer different questions, and an auditor will usually ask about the second.
| Test | Correctly provisioned | Appropriately authorised |
|---|---|---|
| Question it answers | Was access granted according to the rules? | Should this person hold this access today? |
| When it is checked | At the joiner, mover, or leaver event | At defined review intervals, and when risk changes |
| Who confirms it | The provisioning system | A manager, application owner, or entitlement owner |
| What it covers | Access the automation granted | All access the person holds, however it was granted |
| Evidence it produces | A provisioning log | A review record with a decision and a named reviewer |
Identity Governance and Administration (IGA) exists to answer the second question. It does not replace provisioning. It adds the review, validation, and evidence that provisioning cannot supply.
A full IGA programme covering every application, role, and policy is a multi-year undertaking. For many organisations the immediate need is narrower: show that access to a defined set of applications is reviewed, that leavers are removed everywhere, and that exceptions are visible. A focused set of governance controls meets that need:
These controls work best when they start from the applications that an auditor, regulator, or internal risk function will examine, rather than from the whole estate. Each control then produces one evidence view for that scope: what was reviewed, who reviewed it, what changed, and what remains open.
Intragen's right-sized approach to identity governance follows this principle. It extends the identity platforms already in place, connects the applications that create the gap, and puts governance controls in place for a defined scope. The result is a clear record of what is addressed and what is not, with a route to wider IGA maturity when the organisation is ready for it.
A short internal check will show how much of your access risk sits outside the provisioning process. These five questions are a useful starting point:
If more than one answer depends on a spreadsheet, a manual check, or one person's knowledge, that is where governance controls will make the most difference.
Find out where your identity governance gap is. Intragen's identity governance compliance gap assessment is a focused online working session with an identity governance specialist. It identifies the controls you need to evidence, where your current capability falls short, and whether a focused package closes the gap. Book your gap assessment.