Engineer working on a laptop connected to a device test rig, surrounded by wiring harnesses and test hardware

Firmware vulnerabilities & fixes

NIST SP 800-40 Rev. 4: enterprise patch management planning, explained

Last updated:

NIST Special Publication 800-40 Revision 4, "Guide to Enterprise Patch Management Planning: Preventive Maintenance for Technology," defines enterprise patch management as "the process of identifying, prioritizing, acquiring, installing, and verifying the installation of patches, updates, and upgrades throughout an organization." Published in April 2022 by Murugiah Souppaya (NIST) and Karen Scarfone (Scarfone Cybersecurity), it deliberately steps back from patch-management technology — the focus of its 2013 predecessor — to patch-management planning, treating patching as "a cost of doing business" and preventive maintenance for technology rather than a disruption to be tolerated only when forced.

Key takeaways

  • Know when new software vulnerabilities affect your organization's assets (vendor notices, researchers, the NVD).
  • SP 800-40 warns that overly simple patch metrics (e.g.
  • No. The word "maturity" does not appear in Rev.
  • Synthetic example, not a real product or customer.

Four versions, two different jobs

SP 800-40 has been rewritten with a different emphasis each time: the original 2002 edition, Procedures for Handling Security Patches, covered basic patching mechanics; Version 2.0 (2005), Creating a Patch and Vulnerability Management Program, added process and metrics; Revision 3 (2013), Guide to Enterprise Patch Management Technologies, assumed readers already understood the basics and focused on implementing patch-management tooling. Revision 4 deliberately reverses that trajectory: the document states it "is based on the assumption that, in the overall scope of enterprise patch management, organizations would benefit more from rethinking their patch management planning than their patch management technology," and discusses specific products only minimally, pointing instead to the companion NIST SP 1800-31 (also April 2022) for technology implementation detail. Section 1.1 of the SP 800-40 Rev. 4 PDF sets the scope wide: "all types of computing technology assets – including IT, operational technology (OT), Internet of Things (IoT), mobile devices, and cloud computing – and to all types of patchable software – including applications, operating systems, and firmware." Firmware on a device you ship is inside that scope.

Four ways to respond to a vulnerability

Section 2.1 frames patching as one of four possible risk responses to a software vulnerability, not the only one:

Response What it means
Accept "Accept the risk from vulnerable software as is" — rely on existing controls, or judge the potential impact low enough to need no action
Mitigate "Reduce the risk by eliminating the vulnerabilities" (patch, disable the feature, upgrade) "and/or deploying additional security controls" (firewalls, network segmentation)
Transfer "Reduce the risk by sharing some of the consequences with another party" — cybersecurity insurance, or moving to a SaaS provider that owns the patching
Avoid "Ensure that the risk does not occur by eliminating the attack surface" — uninstall the software, decommission the asset, or disable the vulnerable capability

By default, per Section 2.1, "an organization accepts the risk posed by using its software." Patching, updating or upgrading are the only responses, per the document, that can eliminate a vulnerability outright "without removing functionality" — every other response trades away some capability, cost or risk-ownership instead.

What does the patching life cycle look like?

Section 2.2 describes a software vulnerability management life cycle that applies across all four risk responses:

  1. Knowwhen new software vulnerabilities affect your organization's assets (vendor notices, researchers, the NVD).
  2. Planthe risk response: assess the risk, choose a response, decide how to carry it out.
  3. Executethe risk response, in four sub-steps: prepare, implement, verify, and continuously monitor the risk response.

For the patching case, Section 2.3 names the same four sub-steps "Prepare to Deploy the Patch," "Deploy the Patch," "Verify Deployment" and "Monitor the Deployed Patches." Preparing includes a check device makers should design for: "A patch's authenticity and integrity should be confirmed, preferably by automated means, before the patch is tested or installed." Monitoring includes confirming that "the device has not been reset to a vulnerable factory-default state" — a direct hint that a factory reset which rolls back to old firmware undoes the customer's patching.

Four scenarios to plan for, not react to

Section 3.3 sets out four risk-response scenarios every organization should have a maintenance plan for in advance, rather than improvising when one occurs:

Scenario What it covers
Routine patching The standard, regular-release-cycle procedure — most patching, but also the kind most often "postponed and neglected" because it lacks urgency
Emergency patching An accelerated version of routine patching for a severe or actively-exploited vulnerability
Emergency mitigation A temporary measure (disabling functionality, isolating an asset) used before a patch exists, to be replaced by a permanent fix once one is available
Unpatchable assets Long-term isolation or other mitigation for assets that cannot be patched at all, e.g. end-of-life vendor support

Section 3.5.1 says organizations "should consider adopting phased deployments for routine patching in which a small subset of the assets to be patched receive the patch first," with those assets acting as canaries; for emergency patching it "may still be beneficial" to deploy to a few canary assets first. It also suggests giving users some flexibility on routine patches "while also forcing installation after a grace period has ended."

Maintenance groups instead of exceptions

Rather than labeling hard-to-patch assets as one-off "exceptions," SP 800-40 recommends assigning every asset to a maintenance group: a set of assets sharing similar technical and mission characteristics (software to patch, outage restrictions, existing compensating controls, impact if compromised) that therefore need the same maintenance plan for each of the four scenarios above. The document's own worked examples include groups like "legacy OT assets" (no patchable software because it is no longer supported, scheduled outage windows, network isolation and physical security controls as existing mitigations, high impact) and "containers with customer-facing applications in the public cloud" (highly tolerant of downtime, but high impact if compromised) — illustrating that a maintenance group is defined by shared maintenance needs, not just by device type.

Metrics that are actually actionable

SP 800-40 warns that overly simple patch metrics (e.g. a single "percent patched" number) are not actionable because they obscure whether the organization is handling its most important risk first. It illustrates a more useful approach with a notional matrix cross-referencing vulnerability severity against asset importance, reporting the percentage of assets patched by each maintenance plan's deadline plus average and median time-to-patch for each combination — making it possible to see, for example, whether high-severity vulnerabilities on high-importance assets are being remediated meaningfully faster than low-severity ones on low-importance assets, rather than just tracking an aggregate completion rate. In the document's own words, "Overly simplistic metrics, such as counting the number of vulnerabilities that the entire organization had and what percentage of them were patched, are not actionable." The notional Table 1 in Section 3.6 shows the idea with example numbers: critical vulnerabilities on high-importance assets patched by the deadline 95.2% of the time, with a 5.1-day median, against 64.7% and a 75.2-day median for low-importance vulnerabilities on low-importance assets.

Does SP 800-40 include a maturity model?

No. The word "maturity" does not appear in Rev. 4, and it defines no levels or tiers. What it gives instead is a set of four principles in Section 3, which work as a self-check for how far along a patch programme is:

  • "Problems are inevitable; be prepared for them."
  • "Simplify decision making."
  • "Rely on automation."
  • "Start improvements now." — the document adds that some changes "might take years to put in place."

Section 3.6 frames metrics as the improvement mechanism: organizations use them to support "their continuous improvement and effectiveness program" and "to set target values for improving those metrics." If you need a graded maturity score, it has to come from another framework (for example CSF Implementation Tiers or an IEC 62443 maturity level, see our IEC 62443-2-3 patch management guide); SP 800-40 itself measures progress by time-to-patch per maintenance group.

What will your customers ask you before they buy?

SP 800-40 is written for the organization that patches, but Section 3.7, "Consider Software Maintenance in Procurement," turns it into a vendor questionnaire: "Organizations should take software maintenance into consideration when procuring software." Its eight questions are what a device maker should expect from an enterprise or OT buyer that follows NIST:

Section 3.7 question (abridged) What a device maker should have ready
Will you release updates to address vulnerabilities? A written security update policy per product line
How many patches per year, how fast per vulnerability, bundled or not, how often emergency updates? Release cadence and target fix times by severity, based on your own history
"For how many years are you committed to correcting vulnerabilities in the software?" A stated support period and end-of-life date per model
"Will you release updates on a regular schedule, as needed, or both?" A routine release train plus an out-of-band path
Do you have a vulnerability disclosure and incident response program? How transparent are you? A published disclosure policy and advisories (see PSIRT set-up)
Will you provide an emergency mitigation that keeps most functionality? A documented way to disable the affected feature or port
Will the update require restarting the software or rebooting the asset? Reboot behaviour and expected downtime in release notes
How do you test patches, how often do they cause issues, can they be rolled back? Test scope, known-issue history and a rollback mechanism

Section 3.2 adds an inventory point: some vendors "might also provide machine-consumable data on their assets' software composition, such as a software bill of materials (SBOM), which could be used to augment organization inventories." A firmware SBOM per release is how a device maker answers that. On the device side, the Software Update capability in NISTIR 8259A states the matching design goal: "The IoT device's software can be updated by authorized entities only using a secure and configurable mechanism," including the ability to verify an update before install and to roll back. Our NIST IoT device requirements guide covers that baseline, and the firmware OTA update security guide covers the signing and rollback mechanics.

Worked example: how does one firmware fix land in a customer's maintenance groups?

Synthetic example, not a real product or customer. A maker of IP cameras ships firmware 4.2.1 to fix a remote code execution flaw in its web server. A hospital customer runs SP 800-40-style maintenance groups.

  1. KnowThe customer learns of the flaw from the maker's advisory and matches it against the camera models in its inventory, using the SBOM the maker supplies per release.
  2. PlanThe cameras sit in two groups: lobby cameras ("low-impact IoT," routine plan) and cameras in a medication room covered by a safety-related contract ("OT and IoT assets with life-safety impact," which the document lists as an example group). The flaw is in KEV, so the customer chooses emergency patching for both.
  3. PrepareThe customer checks the image's authenticity and integrity (the maker publishes a signature and hash), then tests it on two canary cameras.
  4. ImplementLobby cameras update in the next window. The medication-room cameras are under the manufacturer-controlled schedule Section 2.1 describes, where self-updating "may be voiding the product warranty," so the customer applies the maker's documented emergency mitigation (disable the web server, keep RTSP) until the maker certifies the update.
  5. Verify and monitorThe customer confirms the version on every camera, then re-checks after any factory reset, since a reset to the shipped image would bring the flaw back.

Every step the customer takes depends on something the maker had to publish: the advisory, the SBOM, the signature, the mitigation and the certification status.

What are the common pitfalls?

  • One "percent patched" number. The document calls such metrics "not actionable"; report by maintenance group and importance instead.
  • Treating unsupported devices as exceptions. SP 800-40 wants them in a maintenance group with an "unpatchable assets" plan, not on a waiver list.
  • Periodic scans as the inventory. Section 3.2 says inventorying by monthly or quarterly scans "should no longer be used," and passive scans "will generally underreport vulnerabilities."
  • Skipping integrity checks because the update came from the vendor's own server: the document warns a patch "could have been acquired from a rogue source or tampered with in transit."
  • No end-of-life date. Section 2.1 names the case where "a patch for it will never be released because the software is at end-of-life"; buyers cannot plan for it if the date is unknown.

Where CISA's KEV catalog fits in

SP 800-40 points directly at CISA's Known Exploited Vulnerabilities (KEV) Catalog (CISA also publishes it as machine-readable files in its kev-data repository) as a prioritization input, noting that "CISA's Binding Operational Directive 22-01 requires federal agencies to remediate new vulnerabilities added to the catalog within two weeks," while "other organizations may voluntarily choose to use the catalog to help prioritize their patching efforts." That directive has since been superseded — see our KEV and EPSS prioritization guide for the current BOD 26-04 risk-tiered deadlines and how KEV and FIRST's EPSS, a model that "estimates the probability that a published CVE will be exploited in the wild in the next 30 days," combine into a prioritization signal that feeds directly into the "plan the risk response" step of SP 800-40's life cycle.

Where this sits next to other NIST guidance

Patch management is the execution layer that acts on signals from elsewhere in the NIST stack: it needs the up-to-date asset and vulnerability visibility described in NIST SP 800-137's continuous monitoring (ISCM) guidance to know what to patch in the first place, draws on SP 800-53 controls — Appendix A lists eight Rev. 5 controls as most important for patch management planning (CM-2, CM-3, CM-8, RA-7, SI-2, SR-2, SR-3, SR-5) plus 13 Cybersecurity Framework v1.1 Subcategories, and complements rather than replaces a severity or exploitation-based prioritization layer such as CVSS or SSVC.

A workable starting point

  1. Define your maintenance groups before the next emergency, not during one — group assets by shared patching constraints and impact, not just by device category.
  2. Write a maintenance plan for all four scenarios, including "unpatchable," so a legacy or unsupported asset has a documented mitigation rather than an undocumented gap.
  3. Consider canary rollouts for routine patchingto catch operational problems before they affect the whole fleet, and the same pattern, compressed, for emergency patching.
  4. Track time-to-patch by severity and asset importance together, not as a single aggregate number, so slow remediation of your highest-risk combinations doesn't hide behind a healthy overall percentage.

For a device maker, the first input to all of this is knowing what is inside each firmware image. SoftWSP extracts a CycloneDX SBOM from a firmware image, matches its components against NVD, the CISA KEV catalog, FIRST EPSS and CISA ICS advisories, and drafts a cited report that an engineer reviews before you use it to plan fixes and customer advisories: see device & firmware security services.

This guide summarises a public NIST publication; it does not constitute legal or compliance advice.

Put this guide to work

Frequently asked questions

What is NIST SP 800-40 Rev. 4?

"Guide to Enterprise Patch Management Planning: Preventive Maintenance for Technology," published April 2022 by Murugiah Souppaya (NIST) and Karen Scarfone (Scarfone Cybersecurity). Its own abstract defines enterprise patch management as "the process of identifying, prioritizing, acquiring, installing, and verifying the installation of patches, updates, and upgrades throughout an organization," and frames patching as preventive maintenance for computing technology — "a cost of doing business" rather than an optional chore. It applies to IT, operational technology (OT), IoT, mobile and cloud assets, and to firmware, operating systems and applications alike.

What changed between SP 800-40's earlier versions and Revision 4?

SP 800-40's own introduction gives the lineage: the original SP 800-40, Procedures for Handling Security Patches (2002), covered basic patching procedures; SP 800-40 Version 2.0, Creating a Patch and Vulnerability Management Program (2005), added process, metrics and common-issue content; Revision 3, Guide to Enterprise Patch Management Technologies (2013), assumed readers already understood patching basics and focused on implementing patch management technology. Revision 4 (2022) deliberately shifts focus again — the document states it "is based on the assumption that, in the overall scope of enterprise patch management, organizations would benefit more from rethinking their patch management planning than their patch management technology," and discusses specific tools only minimally.

What are the four risk responses to a software vulnerability, per SP 800-40?

Accept ("accept the risk from vulnerable software as is"), Mitigate ("reduce the risk by eliminating the vulnerabilities" — patching, disabling a feature, upgrading — "and/or deploying additional security controls"), Transfer ("reduce the risk by sharing some of the consequences with another party," e.g. cyber insurance or moving to a SaaS provider who patches on your behalf), and Avoid ("ensure that the risk does not occur by eliminating the attack surface," e.g. uninstalling or decommissioning the vulnerable asset). The document notes that by default, an organization accepts risk simply by continuing to run software, and that patching, updating or upgrading are the only responses that eliminate a vulnerability outright without removing functionality.

What four risk-response scenarios does SP 800-40 say organizations should plan for?

Routine patching (the standard procedure for patches on a regular release cycle, which the document warns is "often postponed and neglected" precisely because it lacks urgency); Emergency patching (an accelerated version of the same process for a severe or actively-exploited vulnerability); Emergency mitigation (a temporary fix — disabling functionality, isolating an asset — used when no patch is yet available, which should later be replaced by a permanent fix); and Unpatchable assets (long-term isolation or other mitigation for assets that cannot be patched, for example because the vendor no longer supports them). SP 800-40 recommends organizations define a maintenance plan for each scenario per maintenance group, rather than treating unpatchable or slow-to-patch assets as ad hoc "exceptions."

How does SP 800-40 relate to CISA's KEV catalog and Binding Operational Directive 22-01?

SP 800-40 cites the CISA Known Exploited Vulnerabilities (KEV) Catalog directly as a prioritization input, and notes that "CISA's Binding Operational Directive 22-01 requires federal agencies to remediate new vulnerabilities added to the catalog within two weeks," while "other organizations may voluntarily choose to use the catalog to help prioritize their patching efforts." See our [KEV and EPSS guide](/guides/kev-epss-vulnerability-prioritization/) for how that directive has since evolved (BOD 22-01 was superseded by BOD 26-04) and how KEV and EPSS combine with a prioritization framework like the one SP 800-40 describes.

Sources

  1. NIST CSRC: SP 800-40 Rev. 4, Guide to Enterprise Patch Management Planning — final publication page
  2. NIST: SP 800-40 Rev. 4, Guide to Enterprise Patch Management Planning: Preventive Maintenance for Technology (full text, April 2022)
  3. CISA: Known Exploited Vulnerabilities Catalog
  4. CISA: Binding Operational Directive 22-01, Reducing the Significant Risk of Known Exploited Vulnerabilities
  5. NIST SP 1800-31: Improving Enterprise Patching for General IT Systems (April 2022)
  6. NISTIR 8259A: IoT Device Cybersecurity Capability Core Baseline (May 2020), Software Update capability (PDF)
  7. CISA: kev-data repository, machine-readable Known Exploited Vulnerabilities catalog files
  8. FIRST: Exploit Prediction Scoring System (EPSS)
Was this guide helpful?

In this cluster

NIST SP 800-40 Rev. 4: enterprise patch management planning, explained