Loading vulnerability…
Trusted-client impersonation in Canonical LXD HTTPS authentication
CVE-2026-112816 is an improper certificate validation vulnerability in the Canonical LXD HTTPS listener, also affecting MicroCloud and Incus. The authentication logic checks every certificate supplied in a client's TLS certificate chain against the server's trust store, although TLS establishes private-key possession only for the leaf certificate. An unauthenticated attacker can append a trusted client's public certificate to a crafted chain and authenticate as that client without possessing its private key. Identified affected LXD versions are 4.0.14, 5.0.10, 5.21.8, and 6.9. Successful exploitation can compromise the cluster, managed virtual machines, and potentially the host.
What it means and what to do now: patch path, interim mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
Authentication bypass allows an attacker to assume a trusted client's identity and exercise its permissions over the cluster and managed virtual machines. Impersonation of a privileged client can provide full administrative access and root access to the host machine, compromising confidentiality, integrity, and availability across the management infrastructure and managed workloads.
If you can’t patch tonight, do this now.
Restrict network access to the HTTPS API to trusted administrative systems until updates are installed. Standalone LXD uses a local socket by default; avoid enabling its network listener where unnecessary. Network restrictions reduce exposure but do not correct the authentication flaw. No configuration-only fix is available.
Patch, then assume compromise.
Upgrade LXD to the fixed release for the installed release line: 4.0.15, 5.0.11, 5.21.9, or 6.10.1. Update MicroCloud to 2.1.5 or 3.4.1 for the corresponding release line, and Incus to 7.6.0. Refresh the applicable LXD and MicroCloud snaps and verify the installed versions afterward, because fixed releases were transitioning from candidate to stable channels when the advisory was issued.
No validated public exploits yet. Hidden: 2 rejected by Mallory's analysis as not a working exploit.
42 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.