Escape Meridian

· Nov 16, 2026

Anthropic: Updating our coordinated vulnerability disclosure policy timelines

Markdown · JSON · Original source

says faster vulnerability discovery by has changed how it handles coordinated vulnerability disclosure (CVD), the process of reporting flaws to maintainers before public release. It is shortening the disclosure window from 90 days to 30 days for high- and critical-severity issues, and to 45 days for all other issues. The new timelines apply to vulnerabilities reported on or after January 1, 2027; reports already in progress keep their existing deadlines. Anthropic says models can help develop fixes and that a bug found by one model is increasingly likely to be found by another. It no longer believes a 90-day window can keep vulnerabilities private for months. Alongside the shorter deadlines, it will escalate reports without a maintainer response after 15 days, rather than 30, and wait 30 days after a patch is available before sharing full technical details, rather than 45. Anthropic says some flaws, including those in hard-to-deploy software or requiring changes to a specification or RFC, may take longer to fix; it will work with vendors on remediation timelines. The seven-day window for vulnerabilities under active exploitation remains unchanged. The update does not apply to AI-discovered findings without human review.

Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.

Two-panel illustration: on the left, a black-and-white photograph of a large boulder, tinted yellow, balanced on smaller rocks; on the right, a blue line drawi…
Anthropic · Two-panel illustration: on the left, a black-and-white photograph of a large boulder, tinted yellow, balanced on smaller rocks; on the right, a blue line drawi… · Open original source

Original source

· Article

Source details

Original source date: Nov 16, 2026 (2026-11-16T00:00:00.000Z).

Date basis: Date stated by the publisher.

Added to Escape Meridian: 2026-11-18T19:48:03.886Z. First observed: 2026-11-17T12:34:02.125Z. Retrieved: 2026-11-17T14:15:46.533Z.

Referenced by

Recorded citations in public readings. A reference alone does not establish agreement or influence.

No public references recorded here yet.

Reference data (JSON)

Limits

does not specify how it will determine remediation timelines for flaws that cannot be fixed within 30 or 45 days; it says it will work with the vendor. The policy update excludes AI-discovered findings that have not received human review, so its new timelines do not cover those findings.

Open question

How will and vendors set remediation timelines for flaws that cannot be fixed within 30 or 45 days?

Source-stated

Anthropic is shortening its CVD disclosure windows to 30 days for high- and critical-severity issues and 45 days for other issues, effective for reports made on or after January 1, 2027.

from 90 days to 30 days for high- and critical-severity issues, and to 45 days for all other issues. The new timelines apply to vulnerabilities reported to vendors and maintainers on or after January 1, 2027.

Source-stated

Anthropic says models can help develop fixes and that bugs found by one model are increasingly likely to be found by another, so it no longer believes a 90-day window can be expected to keep a vulnerability private for months.

Models can now help develop fixes, and a bug that one model can find is increasingly likely to be found by another. We are seeing a higher rate of vulnerability rediscovery as these capabilities increase.

Source-stated

Anthropic will escalate to an external vulnerability coordinator after 15 days without a maintainer response and wait 30 days after a patch is available before sharing full technical details.

we will aim to escalate to an external vulnerability coordinator after 15 days instead of 30 if a maintainer does not respond to our initial vulnerability report. Second, we will wait 30 days instead of 45 after a patch is available

Source-stated

For flaws that cannot be fixed within 30 or 45 days, Anthropic says it will work with the vendor on a remediation timeline; its seven-day disclosure window for vulnerabilities under active exploitation is unchanged.

In these cases, we will work with the vendor on a remediation timeline. Our seven-day disclosure window for any vulnerability under active exploitation remains unchanged.

Source-stated

The updated policy will not apply to AI-discovered findings that have not received human review.

This new policy update will not apply to AI-discovered findings that have not received human review, such as those found by OSS Scanner.

Source-stated

Anthropic says shorter timelines may put more pressure on maintainers and says it will continue providing actionable report guidance and agreeing on a sustainable pace before sending a large volume of findings to one project.

We recognize that shorter timelines may put more pressure on software maintainers. To reduce that burden, we will continue to include actionable guidance with each report

Source and contributor context

Anthropic · Article author

These are credits stated by the source; they do not establish a profile match or authorship of every passage.

Source media

Media declared by the original publisher; availability checked separately from the editorial reading.

Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.

Frontier model

One of the most capable AI models available at a given time; the set shifts as models improve.

Anthropic

Anthropic develops Claude and publishes research about AI capability, safety and how people use its systems.

Editorial provenance

Version 1. Publication ID: b1fb3ac7-3470-42fc-896d-1f27ed3a58af.

Prepared by gpt-6-luna; grounding review by gpt-6-luna.

Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-18T19:48:03.886Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: material_development.

Observation: f5ef54e22bb907fa84a3e6f7a191fd4b9af98d45076eda5dfa3126411b21f98d. Source text hash: 9616703e40b19974e9187d9d536efc1a079ac059fcfbffd86ab75cfd0e1337f5.

Revision and source-check history

Version 1 · 2026-11-18T19:48:03.886Z

Anthropic: Updating our coordinated vulnerability disclosure policy timelines

says faster vulnerability discovery by has changed how it handles coordinated vulnerability disclosure (CVD), the process of reporting flaws to maintainers before public release. It is shortening the disclosure window from 90 days to 30 days for high- and critical-severity issues, and to 45 days for all other issues. The new timelines apply to vulnerabilities reported on or after January 1, 2027; reports already in progress keep their existing deadlines. Anthropic says models can help develop fixes and that a bug found by one model is increasingly likely to be found by another. It no longer believes a 90-day window can keep vulnerabilities private for months. Alongside the shorter deadlines, it will escalate reports without a maintainer response after 15 days, rather than 30, and wait 30 days after a patch is available before sharing full technical details, rather than 45. Anthropic says some flaws, including those in hard-to-deploy software or requiring changes to a specification or RFC, may take longer to fix; it will work with vendors on remediation timelines. The seven-day window for vulnerabilities under active exploitation remains unchanged. The update does not apply to AI-discovered findings without human review.

Original source

· Article

Source details

Original source date: Nov 16, 2026 (2026-11-16T00:00:00.000Z).

Date basis: Date stated by the publisher.

Added to Escape Meridian: 2026-11-18T19:48:03.886Z. First observed: 2026-11-17T12:34:02.125Z. Retrieved: 2026-11-17T14:15:46.533Z.

Limits

does not specify how it will determine remediation timelines for flaws that cannot be fixed within 30 or 45 days; it says it will work with the vendor. The policy update excludes AI-discovered findings that have not received human review, so its new timelines do not cover those findings.

Open question

How will and vendors set remediation timelines for flaws that cannot be fixed within 30 or 45 days?

Source-stated

Anthropic is shortening its CVD disclosure windows to 30 days for high- and critical-severity issues and 45 days for other issues, effective for reports made on or after January 1, 2027.

from 90 days to 30 days for high- and critical-severity issues, and to 45 days for all other issues. The new timelines apply to vulnerabilities reported to vendors and maintainers on or after January 1, 2027.

Source-stated

Anthropic says models can help develop fixes and that bugs found by one model are increasingly likely to be found by another, so it no longer believes a 90-day window can be expected to keep a vulnerability private for months.

Models can now help develop fixes, and a bug that one model can find is increasingly likely to be found by another. We are seeing a higher rate of vulnerability rediscovery as these capabilities increase.

Source-stated

Anthropic will escalate to an external vulnerability coordinator after 15 days without a maintainer response and wait 30 days after a patch is available before sharing full technical details.

we will aim to escalate to an external vulnerability coordinator after 15 days instead of 30 if a maintainer does not respond to our initial vulnerability report. Second, we will wait 30 days instead of 45 after a patch is available

Source-stated

For flaws that cannot be fixed within 30 or 45 days, Anthropic says it will work with the vendor on a remediation timeline; its seven-day disclosure window for vulnerabilities under active exploitation is unchanged.

In these cases, we will work with the vendor on a remediation timeline. Our seven-day disclosure window for any vulnerability under active exploitation remains unchanged.

Source-stated

The updated policy will not apply to AI-discovered findings that have not received human review.

This new policy update will not apply to AI-discovered findings that have not received human review, such as those found by OSS Scanner.

Source-stated

Anthropic says shorter timelines may put more pressure on maintainers and says it will continue providing actionable report guidance and agreeing on a sustainable pace before sending a large volume of findings to one project.

We recognize that shorter timelines may put more pressure on software maintainers. To reduce that burden, we will continue to include actionable guidance with each report

Source and contributor context

Anthropic · Article author

These are credits stated by the source; they do not establish a profile match or authorship of every passage.

Source media

Media declared by the original publisher; availability checked separately from the editorial reading.

Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.

Frontier model

One of the most capable AI models available at a given time; the set shifts as models improve.

Anthropic

Anthropic develops Claude and publishes research about AI capability, safety and how people use its systems.

Editorial provenance

Version 1. Publication ID: b1fb3ac7-3470-42fc-896d-1f27ed3a58af.

Prepared by gpt-6-luna; grounding review by gpt-6-luna.

Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-18T19:48:03.886Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: material_development.

Observation: f5ef54e22bb907fa84a3e6f7a191fd4b9af98d45076eda5dfa3126411b21f98d. Source text hash: 9616703e40b19974e9187d9d536efc1a079ac059fcfbffd86ab75cfd0e1337f5.

uncertain · 2026-11-18T19:48:03.886Z · anthropic/claude-opus-5.5

Publication: b1fb3ac7-3470-42fc-896d-1f27ed3a58af

attribution — pass: The reading consistently attributes the policy changes, rationale and caveats to Anthropic on its own publisher page. Source assertions, such as rising rediscovery rates, remain Anthropic's claims rather than established facts.

We’re shortening the disclosure window in our coordinated vulnerability disclosure (CVD) policy from 90 days to 30 days

summary — uncertain: The core changes are accurate, but the reading leaves out stakes the source states directly. It omits the backlog step: overdue high- and critical-severity reports get 30 more days, then publication. The summary body leaves out Anthropic's acknowledgment of extra pressure on maintainers and its support programs. Smallest repair: add one sentence on each. RFC (a published technical standard document) is unexplained but not central. The reading also says Anthropic will escalate after 15 days, dropping that it will aim to escalate to an external coordinator.

We are now notifying the maintainers of these projects that we are providing them an additional 30 days, after which time we will publish these reports on our dashboard

date — pass: originalPublished, 2026-11-16, matches the publisher page and the feed entry. No earlier first publication is indicated. The January 1, 2027 effective date is stated correctly.

support — uncertain: Most claims are directly supported by their excerpts. Claim 2 recasts the rationale as a 90-day window being unable to keep vulnerabilities private. The source says a vulnerability can no longer be expected to stay private for months, so 90 days no longer makes sense. That conclusion also falls outside the claim's chosen excerpt. Claim 3 states escalation as firm, but the source says Anthropic will aim to escalate. Smallest repairs: restore 'aim to' and quote the private-for-months sentence.

A 90-day window made sense when a vulnerability could be expected to stay private for months. We no longer believe it can

media — pass: The selected media is the verified direct link to the original Anthropic article, which is the most direct and relevant access to the source.

Challenge or correct this reading

Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.