On this page (7 sections)

Google's Project Zero is changing how long it waits before making the bugs it finds public. According to Help Net Security and Infosecurity Magazine, the team will publish technical details 90 days after reporting a bug to a vendor, whether or not a fix has shipped, and will shorten that window to 60 and then 30 days.

This is a policy change, not a single vulnerability, so there is no CVE to patch today. It matters to anyone who runs websites or servers, because details of bugs in common software will appear sooner and the time you have to apply updates will shrink. Note that the research behind this article consists of press reports; the original Project Zero post was not available to us, so the details below are as those outlets describe them.

What Project Zero announced

Since 2021 Project Zero followed a "90 + 30" rule: vendors had up to 90 days to fix a bug, then a further 30 days of "soak time" so the patch could reach users before details were published. The new policy, called Full X, removes the extra month. Details are published X days after the report, where X is 90 today, 60 from February 2, 2027 and 30 from January 3, 2028, as Help Net Security reports.

Full X schedule as reported
PhaseStartsDetails published after
Full 90Now (policy announced November 13, 2026)90 days
Full 60February 2, 202760 days
Full 30January 3, 202830 days

The grace period and the variant policy are also gone. Three exceptions remain, per Infosecurity Magazine: vulnerabilities actively exploited in the wild, bugs whose details are already public, and cases where the vendor asks Project Zero to publish. Help Net Security adds that hardware bugs may still get extra time, as with Spectre and Meltdown.

Why Project Zero says it is doing this

As reported, Project Zero ties the change to large language models, with three arguments:

  • Attackers can find and exploit bugs faster, and the chance rises that several parties, including malicious ones, find the same bug while it is under embargo.
  • The time from patch to working exploit is shrinking. Project Zero now treats zero patch-to-exploitation time as "a reasonable, though imperfect, model of attacker behavior", so waiting a month after a patch gives a false sense of secrecy.
  • Volume has jumped. The team said, "We are finding several vulnerabilities a week in targets where we once found a few vulnerabilities a year."

Severity will not set the timing. Project Zero argues that severity affects the impact on users, not how likely attackers are to discover a bug. Fix status will still be tracked but will no longer decide when details go public.

Vendor concerns and open questions

Before setting the schedule, Project Zero consulted the vendors it reports to most. Their worries, per both reports, were that a sudden switch could expose unfixed bugs, that vendors might fix Project Zero's reports ahead of more severe ones, and that vendors unable to meet an immediate deadline might stop trying. The staged schedule is Project Zero's answer, which it calls "challenging but realistic".

What we cannot say from the research: which vendors will miss deadlines, how many bugs will go public unpatched, or how other researchers will react. Treat any prediction on those points as speculation.

This describes an arrival rate that most patch programmes were never built for.

Ivan Milenkovic, Qualys, to Infosecurity Magazine

What it means for website owners and admins

Project Zero mostly reports bugs in widely used platforms, so the effect reaches you indirectly: browsers, operating systems, libraries and server components you depend on. Realistically, three things change:

  • Public technical write-ups of bugs will appear on a fixed date, even if your vendor has not shipped a fix. You may need mitigations, not just updates.
  • After a patch is released, assume attackers can work out the bug quickly. Qualys's Milenkovic said patching monthly now means "you're slower than disclosure".
  • Vendors may offer fixes that are not patches, such as temporarily disabling a code path. Project Zero expects these to become more common, so read vendor advisories for workarounds as well as versions.
The policy does not change how bugs are fixed, only when details become public. Software you run is not newly vulnerable because of this announcement.

What to do now

  1. List what you run: CMS and plugins, web server, language runtimes, libraries, operating system, router and firewall firmware. You cannot patch what you have not inventoried.
  2. Move critical internet-facing components from monthly patching to a faster routine, for example weekly checks with same-day handling of security releases.
  3. Turn on automatic security updates where they are safe, and subscribe to vendor security announcements for your main software.
  4. Write down a mitigation plan for each key component: how you would disable a feature, restrict access or put a firewall rule in front of it if a fix is not yet available.
  5. Remove software you no longer use. Fewer components means fewer public bug write-ups that apply to you.
  6. Track exposure after disclosure dates: when a vendor advisory or public write-up appears for something you run, check your version the same day.

On Debian or Ubuntu servers, this read-only command lists packages with pending upgrades so you can see how far behind you are. Run it on the server; any package listed has an update available.

List pending package upgrades (Debian/Ubuntu, read-only)Shell
apt list --upgradable

On RHEL-family systems, this lists available security updates without installing anything.

List pending security updates (RHEL/Fedora family, read-only)Shell
dnf updateinfo list security

For WordPress sites, this WP-CLI command shows plugins with updates waiting. Run it from the site directory; entries marked as available need attention.

List plugins with available updates (WordPress, read-only)WP-CLI
wp plugin list --update=available
Tip. Measure your own patch gap: note the date a security update was released and the date it reached production. If the difference is routinely longer than a few days for internet-facing systems, that is the number to reduce.

Detection and monitoring

Shorter disclosure windows mean you should learn about problems from your own monitoring instead of waiting for a monthly review. Watch vendor advisory feeds for your stack, keep an eye on certificate, DNS and header changes that could signal tampering, and run regular outside-in scans of your sites. Webforta Monitor covers uptime, certificate, DNS and header-change monitoring, and the free security headers checker is a quick way to review one layer of hardening. Neither replaces patching.

For background, see our guides on WordPress plugin vulnerabilities and the WordPress security checklist.

Timeline

  • 2021: Project Zero adopts the "90 + 30" policy (per Help Net Security).
  • November 13, 2026: Project Zero outlines the new policy in a post (per Infosecurity Magazine).
  • November 16, 2026: Both outlets report on the change; Full 90 applies.
  • February 2, 2027: Full 60 begins.
  • January 3, 2028: Full 30 begins.

Sources

This article is based on the public sources below. Details can change as vendors and researchers publish more; check the original advisories before acting on version numbers.

  1. Google Project Zero to Cut Disclosure Deadline to 30 Days by 2028infosecurity-magazine.com ·
  2. Google Project Zero will disclose bugs after 90 days, fixed or not - Help Net Securityhelpnetsecurity.com ·
  • Project Zero
  • disclosure policy
  • patch management
  • vulnerability disclosure
  • patch gap