Google's elite vulnerability research unit, Project Zero, has overhauled its long-standing vulnerability disclosure policy, eliminating its traditional grace period and setting a timeline that will reduce vendor remediation windows to just 30 days. As reported by SecurityWeek, the move comes in response to the rise of AI-assisted vulnerability discovery and automated exploitation, which have drastically cut the time between vulnerability identification and active exploitation.

Historically, Project Zero provided software vendors with 90 days to issue a fix after receiving a private bug report, accompanied by an optional 30-day grace period to allow organizations time to roll out patches to end users. That "90 + 30" model is now defunct. Effective November 13, the team has implemented an approach dubbed the "Full X" policy, removing the 30-day roll-out period and establishing rigid disclosure schedules.

Phased Rollout of the 'Full X' Model

Under the new rules, Project Zero will publicly publish technical details of reported vulnerabilities exactly at the deadline mark, regardless of whether a vendor has issued a security patch. The policy is rolling out in three distinct phases:

  • Full 90 (Current): Effective November 13, flaws are disclosed 90 days after reporting, with no additional grace period.
  • Full 60: Beginning February 2, 2027, the disclosure deadline shrinks to 60 days.
  • Full 30: Beginning January 3, 2028, the window contracts permanently to 30 days.

Project Zero security engineer Natalie Silvanovich explained that a gradual transition was selected to prevent vendors from becoming overwhelmed. Software makers consulted during the process raised concerns that an immediate jump to a tight timeline could force them to neglect critical older bugs in favor of newer Project Zero reports, or lead some vendors to abandon remediation deadlines entirely.

The research group also scrapped its previous "soak time" and variant tracking exceptions. As noted in coverage by Infosecurity Magazine, Project Zero observed that soak time—delaying details while users patch—is no longer effective because threat actors routinely reverse-engineer patches almost immediately to extract exploitable flaw mechanics.

Exceptions and AI-Driven Pressures

The team is preserving very few exceptions to the strict deadlines. Exploits actively discovered in the wild will still carry an urgent seven-day disclosure timeline, and bugs already disclosed publicly by outside researchers will not be delayed. Vendors may also request earlier disclosure if their patch is ready ahead of the deadline. Historical edge cases involving sweeping architectural flaws, such as Spectre and Meltdown, will continue to receive rare, coordinated exceptions.

Underpinning the policy shift is an explosion in vulnerability research speed enabled by automation and machine learning. Project Zero researchers stated that they now identify several security flaws a week in targets where they previously uncovered only a few per year. The group stressed that an issue's severity should not dictate disclosure timing, observing that high severity influences user impact rather than the likelihood that an adversary will discover the flaw.

Security industry leaders point out that enterprise patch workflows are ill-prepared for this pace. Ivan Milenkovic, vice president of risk technology at Qualys, told Infosecurity Magazine that Project Zero has effectively "removed the snooze button," adding that monthly patching cadences are now outpaced by public disclosure timelines.

What Defenders Should Do

Security teams and enterprise IT departments must adapt to faster disclosure lifecycles as researchers and threat actors accelerate bug discovery. Defenders should evaluate the following operational adjustments:

  • Shorten internal patch cycles: Re-evaluate traditional 30-day or monthly patch schedules, particularly for internet-facing systems, to align with public disclosures where exploit details may appear without warning.
  • Prioritize compensating controls: Because technical vulnerability details will regularly become public before downstream enterprise deployments are complete, establish repeatable virtual patching mechanisms using intrusion prevention systems (IPS) and web application firewalls (WAF).
  • Track vendor disclosure posture: Monitor which third-party software providers struggle to meet the 90-day and upcoming 60-day deadlines, factoring their patch velocity into third-party supply chain risk assessments.
  • Automate patch validation: Streamline testing pipelines for operating systems and business-critical software to reduce the time spent approving security updates once released.