Project Zero plans to shorten its standard disclosure window from 90 days to 30 by 2028, citing faster AI-assisted exploit development. Is that enough time for vendors to produce good fixes, and what does it mean for teams applying patches?
Is a 30-day vulnerability disclosure window realistic?
Answers
The change puts pressure on two different processes: vendors have to build and test a fix, then customers have to deploy it. A shorter disclosure deadline doesn't automatically make either process faster.
Teams whose patching schedules assume a long runway may need a way to assess urgent issues outside the normal release cycle. That could mean identifying affected systems quickly, testing fixes sooner and having temporary mitigations ready when a patch isn't available.
Reply to this answer
Thirty days can be tight for a complicated fix. Rushing changes through development and testing risks a patch that breaks things or doesn't fully address the vulnerability. Disclosure without a fix can leave users in a rough spot too.
I'd want vendors to communicate useful mitigations early, rather than treating the deadline as the only thing that matters.
Reply to this answer
The argument for a shorter window is that attackers may be able to develop exploits faster with AI assistance. If that happens, a long private window isn't necessarily safe just because the details haven't been published.
Some people would shorten it further, even to a week. But claims that an AI can reliably turn any bug into a working exploit in minutes are a lot stronger than the evidence in this discussion supports.
_g53!mivquaro.7:
That distinction helps. Faster exploitation is a reason to revisit the timeline, but not proof that every vulnerability can be exploited immediately.
Reply to this answer
This is also an argument for continuous exposure management, not just tracking patch deadlines. Keep an eye on which vulnerabilities affect reachable systems and what can be mitigated while a fix is being developed or tested.
Automation may help teams respond faster, but it doesn't remove the need to check whether a fix works and what it might disrupt.
Reply to this answer
Add your answer
Answers are reviewed before publication.
6h!fablebadge_293:
That also makes the distinction between a vendor releasing a patch and customers actually installing it pretty important.