Updating our coordinated vulnerability disclosure policy timelines
According to Anthropic Research: The Frontier Red Team reports that AI models are getting much better at discovering software vulnerabilities, increasing the risk of independent rediscovery.

Original image temporarily unavailable
According to Anthropic Research:
The Frontier Red Team states that because advanced AI models can now discover software vulnerabilities and help develop fixes at unprecedented speeds, the traditional 90-day coordinated vulnerability disclosure (CVD) window is no longer viable, as bugs are likely to be rediscovered independently by other systems.
Starting January 1, 2027, the disclosure window for high- and critical-severity issues will be shortened to 30 days, while all other issues will drop to 45 days. Additionally, if software maintainers do not respond within 15 days, the team will escalate the issue to an external vulnerability coordinator, and full technical details will be shared 30 days after a patch becomes available.
While active exploitation cases retain a seven-day window and complex software with hard-to-deploy fixes can negotiate custom timelines, the team is offering support programs like OSS Scanner and free Claude access to help maintainers cope with the accelerated pace.
View selection rationale
Importance 75/100
About this brief
- Content type
- News summary
- Source type
- First-party / official source
- Original source
- Anthropic Research ↗
- Source published
- Process
- AI-assisted discovery, translation, summarisation, classification and scoring; standards and corrections are maintained by the editorial desk.
- Verification
- The original URL and publication metadata are retained. This brief is not independently fact-checked unless explicitly stated.
- Published by
- Polyemergence Editorial Desk · PolyEmergence Pte Ltd.
- Conflict disclosure
- No financial or commercial conflict declared.