LXD server lets attackers bypass login and take over the cluster
A flaw in LXD, a tool for running virtual machines and containers on servers, lets an outsider pretend to be a trusted admin. This can give them control of the whole cluster and every virtual machine on it.
| Severity | CriticalCVSS 10 of 10 |
|---|---|
| Being exploited? | UnclearNo reports either way yet |
| Fix available? | YesLXD 4.0.15, 5.0.11, 5.21.9, 6.10.1 or later · since 30 Oct 2026 |
| How sure is this? | Official, 2 sourcesBacked by an official source |
How the attack works simplified
- 1Outside attackerNo login or account needed
- 2Crafted certificateForged chain sent to HTTPS listener
- 3LXD accepts itCertificate validation is improper
- 4Trusted client accessAttacker is treated as a trusted admin
- ResultAttacker controls the cluster and its VMs
What happened
Improper certificate validation in the HTTPS listener of Canonical LXD allows an unauthenticated attacker to authenticate as any trusted client using a crafted certificate chain.
Are you affected?
You are affected if you run LXD 4.0.14 or earlier 4.0.x builds, 5.0.10 or earlier 5.0.x builds, 5.21.8 or earlier 5.21.x builds, or 6.9 with an HTTPS listener reachable by others.
LXD 4.0.14 (and earlier 4.0.x); LXD 5.0.10 (and earlier 5.0.x); LXD 5.21.8 (and earlier 5.21.x); LXD 6.9
What to do
- Update LXD to 4.0.15, 5.0.11, 5.21.9, or 6.10.1 or later on every server and cluster member
- Restrict network access to the LXD HTTPS port so only trusted management hosts can reach it
- Review LXD trust store and access logs for unexpected client certificates
Timeline
- 30 Oct 2026, 17:17 UTCPublicly disclosed
- 30 Oct 2026, 17:17 UTCReported by NVD · NVD
- 30 Oct 2026, 17:19 UTCPicked up by Zero Radar
- 30 Oct 2026, 23:02 UTCHacker News reported on it · Hacker News
Sources 2 reports, oldest first
Official vendor, government CERT or advisory database
- CVE-2026-112816NVD · 30 Oct 2026, 17:17 UTC
Community discussion sites, treat with care
- CVE 10.0: CVE-2026-112816, no emudhra requiredHacker News · 30 Oct 2026, 22:26 UTC
Technical details
- ID
- CVE-2026-112816
- CVSS
- 10 / 10
- Disclosed
- 30 Oct 2026, 17:17 UTC
- On Zero Radar
- 30 Oct 2026, 17:19 UTC (2 min after disclosure)
- Last change
- 30 Oct 2026, 23:02 UTC
The plain-language parts and the attack diagram were written by an AI model (anthropic/claude-haiku-5.5) from the sources above. Facts like versions, exploitation and dates come from the sources themselves. How Zero Radar works