Home / Blog / Updating our coordinated vulnerability disclosure policy…
Engineering

Updating our coordinated vulnerability disclosure policy timelines

Anthropic shortened its coordinated vulnerability disclosure timeline from 90 days to 30 days for critical flaws, taking effect starting January 1, 2027.

By Dillip Chowdary • Nov 16, 2026 • Source: Anthropic Research

Updating our coordinated vulnerability disclosure policy timelines
Image: Anthropic Research

At a glance

Effective date
January 1, 2027
Critical flaw window
30 days
General flaw window
45 days
Coordinator escalation
15 days
Claude Ultra grant
Six months at no cost
Exploited bug window
7 days

Anthropic announced major updates to its coordinated vulnerability disclosure policy timelines, reducing the resolution window for high-severity and critical-severity bugs from 90 days to 30 days. According to details outlined in Anthropic Research's report, the disclosure window for all other vulnerabilities will drop to 45 days. The updated policy takes effect for issues reported to vendors and project maintainers on or after January 1, 2027.

This analysis examines the condensed disclosure deadlines, new escalation triggers for non-responsive vendors, and the impact of automated vulnerability discovery on defensive security. It also details remediation exceptions for critical infrastructure, volume pacing protections for maintainers, and dedicated defensive grants. Security teams, software vendors, and open-source maintainers must adapt their incident response workflows to meet these shorter timelines.

01Revised deadlines for vulnerability disclosure

Anthropic is shortening its coordinated vulnerability disclosure timelines to match the speed of modern security research. Starting January 1, 2027, high-severity and critical-severity vulnerabilities must be remediated within 30 days before public disclosure, while all other issues receive a 45-day window. In-progress vulnerability reports filed prior to January 1, 2027, will maintain their existing deadlines under the previous policy.

| Disclosure Policy Metric | Previous Window | Updated Window | | High and critical severity | 90 days | 30 days | | All other vulnerability severities | 90 days | 45 days | | Escalation for non-responsive maintainers | 30 days | 15 days | | Public technical details post-patch | 45 days | 30 days | | Actively exploited vulnerabilities | 7 days | 7 days |

Two related operational milestones will also accelerate under the revised disclosure policy. If a project maintainer fails to respond to an initial report, Anthropic will escalate the issue to an external vulnerability coordinator after 15 days instead of 30 days. Additionally, Anthropic will wait 30 days instead of 45 days after a vendor patch becomes available before publishing complete technical details.

Maintainers responsible for existing unresolved issues also face an impending disclosure deadline under the transition terms. Anthropic identified critical and high-severity reports submitted over 90 days ago that remain undisclosed past current policy limits. The organization is providing affected maintainers an additional 30 days of notice before publishing these vulnerability reports on its public dashboard starting January 1, 2027.

02Frontier models and automated bug rediscovery

Established vulnerability disclosure norms were designed when software security research moved exclusively at human speed. Historical policies assumed that crafting patches required lengthy manual deliberation and that rival researchers were unlikely to discover the same underlying flaw independently. Modern frontier models have invalidated both assumptions by accelerating remediation assistance and making rapid, independent rediscovery of identical vulnerabilities common.

Anthropic observed higher rates of bug rediscovery as artificial intelligence capabilities advanced over recent months. Because a vulnerability discovered by one model is increasingly likely to be found independently by another system, defensive obscurity degrades rapidly. While a 90-day embargo functioned when vulnerabilities stayed secret for months, lengthy confidentiality windows no longer protect software ecosystems effectively.

The shortened 30-day and 45-day disclosure windows are intended to keep pace with improving artificial intelligence capabilities without overwhelming defenders. Faster automated discovery helps defensive teams identify code defects earlier, but defenders must deploy fixes before external actors uncover identical bugs. Shortened disclosure targets maintain this balance by compelling prompt patch deployment across public and private codebases.

Advertisement

Tech Pulse Daily

Get tomorrow's pulse first

Join engineers who read Tech Pulse before stand-up. Free, weekday mornings.

03Remediation adjustments for specifications and infrastructure

Certain categories of complex software cannot conform to accelerated disclosure schedules without risking widespread instability. Vulnerabilities affecting hard-to-deploy systems, such as critical infrastructure, or flaws requiring formal updates to an internet standard or RFC specification often demand extended remediation. In these scenarios, Anthropic will coordinate directly with affected vendors to formulate customized remediation schedules.

Note: Vulnerabilities in critical infrastructure or RFC specifications operate under customized remediation plans, while actively exploited flaws retain a strict seven-day disclosure window. Automated findings from OSS Scanner lacking human review remain excluded from these policy timelines.

Human review remains a mandatory gate before any artificial intelligence finding enters the coordinated disclosure timeline. The updated 30-day and 45-day timelines do not apply to raw, automated discoveries that have not undergone human analytical verification. This boundary ensures software maintainers are not burdened by unvalidated automated outputs generated during wide-scale scanning routines.

04Mitigation guidance and sustainable delivery pacing

Tighter disclosure windows create operational burdens for open-source maintainers and corporate software engineering teams. To alleviate this strain, Anthropic will accompany each vulnerability notification with actionable guidance, including functional patches or mitigation steps whenever possible. Providing ready-to-test fixes helps maintainers verify reports and resolve identified defects without delay.

Security research teams must also regulate reporting volume to prevent maintainer burnout. When auditing large software projects, Anthropic will negotiate a sustainable reporting pace with maintainers before delivering large batches of vulnerabilities. Establishing mutual pacing agreements prevents small developer teams from being overwhelmed by simultaneous remediation clocks.

05Defensive scanning services and compute grants

Anthropic launched specialized support programs to assist defenders and open-source projects operating under compressed timelines. Maintainers of critical open-source repositories can enroll in OSS Scanner, a no-cost defensive service that periodically scans codebases with Anthropic's strongest frontier models. The scanning platform delivers discovered security findings directly to authorized maintainers for verification and remediation.

Maintainers and active contributors can also receive advanced artificial intelligence tooling through the Claude for Open Source Program. Accepted applicants receive six months of Claude Ultra at no cost to assist with code audits, issue triage, and patch authoring. Providing model access helps open-source stewards match the discovery capabilities of commercial research teams.

Security professionals and open-source maintainers can also join the Cyber Verification Program. This initiative tailors standard safeguards in Claude to facilitate legitimate defensive engineering, vulnerability verification, and patch development. These defensive programs work in tandem to ensure software defenders possess adequate resources to meet shortened disclosure windows.

Frequently asked questions

When do Anthropic's updated vulnerability disclosure timelines take effect?

The updated timelines apply to vulnerabilities reported to maintainers and vendors on or after January 1, 2027. Existing in-progress reports submitted prior to that date will retain their original deadlines.

What are the revised disclosure deadlines for reported vulnerabilities?

The disclosure window is reduced from 90 days to 30 days for high- and critical-severity vulnerabilities, and to 45 days for all other issues. Vulnerabilities under active exploitation retain an unchanged seven-day disclosure window.

How quickly does Anthropic escalate unresponsive vulnerability reports?

Anthropic will aim to escalate a report to an external vulnerability coordinator after 15 days instead of waiting 30 days if a maintainer fails to respond.

What resources are available to help open-source maintainers fix flaws?

Maintainers can use OSS Scanner for free automated repository scans, apply for six months of free Claude Ultra access, and join the Cyber Verification Program for defensive security work.

Sources

Dillip Chowdary

Author

Dillip Chowdary

Writes Tech Bytes coverage of AI, engineering, and the tools that actually ship. Editor of Tech Pulse Daily.

Related on Tech Bytes

Advertisement

5-min tech signal

Weekday briefing for engineers who skip the noise.

No spam · Unsubscribe anytime

Advertisement

✈️ CareerPilot

Your AI job-search copilot

Match your resume against live Ashby, Greenhouse & Lever openings — fit scores, job-specific resume optimization and email alerts.

Find matching jobs →

Free Tools

Browse all tools →