Patch Management Policy

Enterprise patch management policy

Large estates need windows that depend on the asset group, not one number. This variant, modeled on NIST SP 800-40 Rev. 4 (April 2022), assigns every asset to one maintenance group, shortens windows for high importance groups to 7 and 14 days, and lets group owners delegate approval in writing.

Enterprise variant: paste over the same-numbered sections of the general policy

2 Roles and responsibilities

2.1 Policy owner ([CISO]): owns this policy and the windows. Approves changes to maintenance plans.

2.2 Patch program lead ([role]): runs the cycle across all groups and reports monthly to the [risk committee].

2.3 Group owners ([role per maintenance group]): own the maintenance plan for their group, approve outage time, and accept test results. Group owners may delegate day-to-day approval in writing to [role]; the delegation names the group, the limits, and an expiry date no later than [12] months.

2.4 Business unit heads: receive the group-level report and own late items in their units.

2.5 Change advisory board ([name]): reviews only changes that fall outside a maintenance plan. Plan-conforming patches are pre-approved.

3 Asset inventory

3.1 The inventory may be held in more than one system. The patch program lead keeps a register that names each source of record, the asset types it covers, and how often it syncs.

3.2 Each asset is assigned to exactly one maintenance group. Assets that cannot meet their group's plan move to a separate group with its own plan; they are not left as exceptions in the group.

3.3 Authenticated scanning or agent reporting is the source for installed software. Unauthenticated scans are used only for discovery.

4 Severity and patch windows

4.1 Group importance is High, Moderate or Low, set by the group owner and confirmed by the policy owner.

4.2 Patch windows

Windows (calendar days from vendor release)

SeverityWindow3143090180 days
1. Known exploited, internet-facingHigh importance group: 3; Moderate importance group: 3; Low importance group: 3
1. Known exploited, otherHigh importance group: 7; Moderate importance group: 14; Low importance group: 14
2. CriticalHigh importance group: 7; Moderate importance group: 14; Low importance group: 14
3. HighHigh importance group: 14; Moderate importance group: 30; Low importance group: 30
4. MediumHigh importance group: 60; Moderate importance group: 90; Low importance group: 90
5. LowHigh importance group: 180; Moderate importance group: 180; Low importance group: 180

Windows as this template sets them, on the same 0 to 180 day scale as the builder.

5 Testing and staging (rings)

5.1 Ring 0: [IT and security staff], first, for [3] working days.

5.2 Ring 1: [volunteer early adopters across business units], for up to [14] days for routine updates.

5.3 Ring 2 and later: remaining assets, in groups of [size].

5.4 Class 1 updates skip Ring 1 and run Ring 0 for no more than [24] hours.

8 Reporting and metrics

8.1 Reports are cut by importance of asset and class of vulnerability, not as one overall percentage. Each cell shows the share patched by deadline, the mean days and the median days.

8.2 The CISO sees all cells; group owners see their group; the board sees the cells for High importance assets and class 1 and 2 vulnerabilities.

Why maintenance groups replace one policy-wide rule

NIST defines a maintenance group as a set of assets with similar characteristics that generally have the same software maintenance needs for each risk response scenario (section 3.4). It then says organizations should define a maintenance plan for each group and each scenario, including timeframes for beginning and ending each action (section 3.5). The publication sets no day counts of its own. The numbers in the template's table are therefore this site's suggested defaults: the High importance column is set shorter than the general policy, and the other columns equal it.

NIST also says maintenance groups should be defined to minimize assets treated as exceptions, and that assets with similar long-term exceptions might need a separate group with its own plan. Section 3.2 of the template adopts that as a rule: an asset that cannot meet its group's plan moves, it does not sit as an exception.

Inventory without a configuration database

NIST says the software inventory could be in a single enterprise asset inventory or split among multiple resources, that a realistic goal is a close-to-comprehensive inventory built on automation, and that updating inventories monthly or quarterly through a vulnerability scan is a model that should no longer be used. Authenticated scanning is needed if scans supply software inventory. Template section 3.1 therefore asks for a register of the sources of record instead of demanding a CMDB, and 3.3 limits unauthenticated scans to discovery.

Rings and delegation

The ring lengths come from NIST section 3.5.1, which describes a small first round of technically knowledgeable users lasting a few days, then a larger round of early adopters across the organization who try app updates for a few weeks. NIST adds that routine patches should be installable with some flexibility but forced after a grace period, either by direct execution or by blocking network access of unpatched assets. Delegation, the change advisory board carve-out and the sign-off expiry are the site's drafting; no source read sets them.

Where the estate includes financial entities in the EU, DORA's RTS Article 10(4)(d) separately asks for set patch deadlines and escalation procedures when deadlines cannot be met; template section 4 and Annex A carry both (source).

Where to go next

The metrics table in clause 8.1 follows the matrix explained in reporting and metrics. For sign-off wording see roles and responsibilities. Sections not shown above are as in the general policy.

Sources

Last reviewed 19 Nov 2026