The short answer. “Patched” is a verifiable state, not a status field. Confirm the May 2026 cumulative update is installed and rebooted on every in-scope asset, reconcile that against an out-of-band source (vulnerability scanner, Intune/Update for Business report, or PowerShell hotfix query), monitor the CISA Known Exploited Vulnerabilities catalog daily, and document every exception with a compensating control and a remediation date.

Published · HI Tech Hui · ~5 min read

Why the day 3–7 window matters

The first 48 hours after Patch Tuesday is about deployment. The next several days are about verification — the period where most teams find out which assets did not actually take the update, which reboots were skipped, and which exceptions never got written down. Our earlier brief on the May 2026 Patch Tuesday priority list covered what to deploy first, including CVE-2026-41089, the critical Windows Netlogon remote code execution vulnerability affecting domain controllers. This post is the follow-up: how to know, in writing, that the work landed.

The CISA BOD 22-01 framework gives federal civilian agencies two weeks to remediate listed CVEs assigned in 2021 or later. That two-week clock is the right default for any organization treating KEV as an operational input. By day 7, you should have evidence in hand — not assurances.

Out-of-band verification beats dashboard verification

Every patch management tool reports its own state. WSUS, ConfigMgr, Intune, Action1, Kaseya, NinjaOne, Tanium — they all draw their compliance numbers from the same agent or scan source. If that source is wrong (stale heartbeat, broken WMI provider, misconfigured policy ring), the dashboard is also wrong. Out-of-band verification means asking a different question, against a different source, and reconciling the answers.

Concretely, for the May 2026 cumulative update on Windows Server and Windows 11:

  • Pull the installed update history directly from each host. On a Windows host, Get-HotFix | Sort-Object InstalledOn -Descending | Select-Object -First 10 returns the most recently installed KBs and their install dates. On a fleet, run the same query through your endpoint management tool's remote-script capability or via PowerShell remoting, then compare results to the May 2026 KB numbers published in the Microsoft Update Catalog for that build.
  • Re-scan with a vulnerability scanner that has been refreshed for the May 2026 release. Confirm the scanner’s plugin or detection set was updated to recognize the May 2026 KBs before you trust its results. A scanner that reports “no vulnerabilities” because it has not yet learned about the new CVEs is worse than no scan.
  • Reconcile reboot state. An installed KB that has not been rebooted into is not effective. Query (Get-WmiObject -Class Win32_OperatingSystem).LastBootUpTime and confirm the host has rebooted after the install time of the May 2026 KB.
  • Cross-check identity and edge assets a second time. Domain controllers, internet-facing servers, and identity infrastructure should be checked by at least two independent methods — for example, an EDR-side query plus a patch-tool report — before they are marked verified.

Track the KEV catalog this week, not next month

At release, Microsoft listed CVE-2026-41089 as not publicly disclosed, not exploited, and with exploit code maturity unproven. That is the picture at release, not a forecast. The interesting question is what changes during the first three weeks after disclosure — the period when proof-of-concept code, weaponized tooling, and confirmed exploitation in the wild typically appear for high-impact Netlogon and identity bugs.

For day 3–7 operations, the practical move is to put the CISA KEV catalog on a daily checklist for the month after a critical Patch Tuesday. CISA adds entries when it has reliable evidence of in-the-wild exploitation; once added, federal civilian agencies inherit BOD 22-01 timelines, and the default catalog timeline for post-2020 CVEs is two weeks. Private-sector teams in Hawaii and elsewhere should adopt the same window as a defensible baseline. If a CVE you have not yet remediated lands on KEV, the right reaction is an emergency change, not a wait for next month’s window.

Evidence that holds up

If you have to defend the patch cycle to a regulator, an insurer, a board, or your own future self after an incident, “the dashboard said green” is not enough. For each in-scope asset class, keep the following on file for the May 2026 release:

  • The installed KB number and install timestamp per host (from a query, not a screenshot of the patch tool).
  • The post-install reboot event — either the boot time after install, or an explicit pending-reboot resolution log.
  • A vulnerability scanner result, re-baselined against the May 2026 release, showing the host clean for the May 2026 CVEs you care about.
  • For domain controllers specifically: confirmation that the Netlogon and LSA services restarted cleanly and that authentication telemetry looks normal in the days after the patch window.
  • An exception register for any asset not patched in the window, with owner, business justification, compensating control, and remediation date.

Handling the exceptions you actually have

Every real environment has exceptions. The point of a written exception is not to make them go away; it is to make them visible, time-bound, and assigned. For the May 2026 cycle, the exceptions we see most often on Hawaii networks are:

  • Domain controllers in change-frozen environments. Document the freeze, the risk, and the conditions under which it lifts. A frozen DC running a known-critical unpatched CVE needs network segmentation and enhanced monitoring while the freeze is in effect.
  • Legacy line-of-business servers that cannot reboot during business hours. Schedule a real reboot window, in writing, within the BOD 22-01 default of two weeks. “Next maintenance window” without a date is not an exception, it is a delay.
  • Vendor-managed appliances on Windows underneath. Confirm in writing with the vendor that the May 2026 updates are tested and applied, and capture the response. Vendor silence is not patching.

The day 7 close-out checklist

  • Every domain controller is verified patched and rebooted, by two independent methods.
  • Internet-facing Windows servers and identity infrastructure are verified patched and rebooted.
  • Workstation patch compliance is reported and trending toward 100% on the standard ring schedule.
  • The CISA KEV catalog has been checked daily since May 12, and any KEV additions touching this release have been triaged.
  • Exceptions are documented with owner, compensating control, and remediation date.
  • The next patch window (June 2026 Patch Tuesday) is on the calendar with change approvals in flight.

What this looks like for HI Tech Hui clients

For organizations on HI Tech Hui managed IT services, the verification cycle is part of the standard monthly patch program: out-of-band installation evidence per host, reconciliation against the vulnerability scanner, KEV monitoring through the 24/7 Cyberuptive SOC, and documented exception handling. For organizations in regulated industries — healthcare, finance and professional services, government contractors, and law firms — the evidence package is structured against the relevant control families so it can be handed to auditors without rework.

Sources


Need a verification pass on this month’s patch cycle on Hawaii infrastructure? HI Tech Hui’s managed IT services and cybersecurity teams handle out-of-band patch verification, KEV monitoring, and exception tracking, with 24/7 coverage through our SOC. Get in touch.

Ready when you are

Let’s scope your IT & security plan.

Talk with a Honolulu-based engineer about managed IT, cybersecurity, or a 24/7 SOC handoff. We’ll review your current environment, identify the highest-impact gaps, and outline a clear next step — with no obligation.

HI Tech Hui team