TS-2026-014
Vulnerability from tailscale - Published: Mon, 26 Oct 2026 00:00:00 GMT
Description: Tailscale SSH Unix socket forwarding could grant peers access to root-only sockets.
What happened?
Tailscale SSH allows the forwarding of TCP and Unix sockets between a client and a host. The SSH server in the root-privileged tailscaled process validated Unix socket paths lexically but would perform the operations as root. This created three vulnerabilities. First, it allowed privileged dials via local forwarding (ssh -L). Second, it allowed socket replacement via remote forwarding (ssh -R) by unlinking root-owned sockets. Third, a race condition defeated the path allowlist because the kernel re-resolves paths after the initial check.
These issues combined to remove the path allowlist as a security boundary. This gave authenticated peers with authorized forwarding access a route to root-only Unix sockets on the node.
Tailscale now performs Unix socket forwarding on behalf of the authenticated user rather than as root.
This vulnerability is fixed in Tailscale version 1.104.4 or newer.
What was the impact?
An attacker could reach root-only sockets such as the Docker socket or the tailscaled local API.
Who was affected?
This issue affects users with Tailscale SSH enabled who also allow local or remote forwarding in their access policies.
What do I need to do?
If you use Tailscale SSH with local or remote forwarding, upgrade to Tailscale version 1.104.4 or newer.
As a mitigation, you can disable SSH forwarding by setting TS_SSH_DISABLE_FORWARDING=1 in the tailscaled service environment and restarting the service. This disables forwarding without disabling all of Tailscale SSH.
{
"guidislink": false,
"id": "https://tailscale.com/security-bulletins/#ts-2026-014",
"link": "https://tailscale.com/security-bulletins/#ts-2026-014",
"links": [
{
"href": "https://tailscale.com/security-bulletins/#ts-2026-014",
"rel": "alternate",
"type": "text/html"
}
],
"published": "Mon, 26 Oct 2026 00:00:00 GMT",
"summary": "\u003cp\u003e\u003cstrong\u003e\u003cem\u003eDescription\u003c/em\u003e\u003c/strong\u003e: Tailscale SSH Unix socket forwarding could grant peers access to root-only sockets.\u003c/p\u003e\n\u003ch4\u003eWhat happened?\u003c/h4\u003e\n\u003cp\u003e\u003ca href=\"https://tailscale.com/docs/features/tailscale-ssh\"\u003eTailscale SSH\u003c/a\u003e allows the forwarding of TCP and Unix sockets between a client and a host. The SSH server in the root-privileged \u003ccode\u003etailscaled\u003c/code\u003e process validated Unix socket paths lexically but would perform the operations as root. This created three vulnerabilities. First, it allowed privileged dials via local forwarding (\u003ccode\u003essh -L\u003c/code\u003e). Second, it allowed socket replacement via remote forwarding (\u003ccode\u003essh -R\u003c/code\u003e) by unlinking root-owned sockets. Third, a race condition defeated the path allowlist because the kernel re-resolves paths after the initial check.\u003c/p\u003e\n\u003cp\u003eThese issues combined to remove the path allowlist as a security boundary. This gave authenticated peers with authorized forwarding access a route to root-only Unix sockets on the node.\u003c/p\u003e\n\u003cp\u003eTailscale now performs Unix socket forwarding on behalf of the authenticated user rather than as root.\u003c/p\u003e\n\u003cp\u003eThis vulnerability is fixed in Tailscale version 1.104.4 or newer.\u003c/p\u003e\n\u003ch4\u003eWhat was the impact?\u003c/h4\u003e\n\u003cp\u003eAn attacker could reach root-only sockets such as the Docker socket or the \u003ccode\u003etailscaled\u003c/code\u003e local API.\u003c/p\u003e\n\u003ch4\u003eWho was affected?\u003c/h4\u003e\n\u003cp\u003eThis issue affects users with Tailscale SSH enabled who also allow local or remote forwarding in their access policies.\u003c/p\u003e\n\u003ch4\u003eWhat do I need to do?\u003c/h4\u003e\n\u003cp\u003eIf you use Tailscale SSH with local or remote forwarding, upgrade to Tailscale version 1.104.4 or newer.\u003c/p\u003e\n\u003cp\u003eAs a mitigation, you can disable SSH forwarding by setting \u003ccode\u003eTS_SSH_DISABLE_FORWARDING=1\u003c/code\u003e in the \u003ccode\u003etailscaled\u003c/code\u003e service environment and restarting the service. This disables forwarding without disabling all of Tailscale SSH.\u003c/p\u003e",
"summary_detail": {
"base": "https://tailscale.com/security-bulletins/index.xml",
"language": null,
"type": "text/html",
"value": "\u003cp\u003e\u003cstrong\u003e\u003cem\u003eDescription\u003c/em\u003e\u003c/strong\u003e: Tailscale SSH Unix socket forwarding could grant peers access to root-only sockets.\u003c/p\u003e\n\u003ch4\u003eWhat happened?\u003c/h4\u003e\n\u003cp\u003e\u003ca href=\"https://tailscale.com/docs/features/tailscale-ssh\"\u003eTailscale SSH\u003c/a\u003e allows the forwarding of TCP and Unix sockets between a client and a host. The SSH server in the root-privileged \u003ccode\u003etailscaled\u003c/code\u003e process validated Unix socket paths lexically but would perform the operations as root. This created three vulnerabilities. First, it allowed privileged dials via local forwarding (\u003ccode\u003essh -L\u003c/code\u003e). Second, it allowed socket replacement via remote forwarding (\u003ccode\u003essh -R\u003c/code\u003e) by unlinking root-owned sockets. Third, a race condition defeated the path allowlist because the kernel re-resolves paths after the initial check.\u003c/p\u003e\n\u003cp\u003eThese issues combined to remove the path allowlist as a security boundary. This gave authenticated peers with authorized forwarding access a route to root-only Unix sockets on the node.\u003c/p\u003e\n\u003cp\u003eTailscale now performs Unix socket forwarding on behalf of the authenticated user rather than as root.\u003c/p\u003e\n\u003cp\u003eThis vulnerability is fixed in Tailscale version 1.104.4 or newer.\u003c/p\u003e\n\u003ch4\u003eWhat was the impact?\u003c/h4\u003e\n\u003cp\u003eAn attacker could reach root-only sockets such as the Docker socket or the \u003ccode\u003etailscaled\u003c/code\u003e local API.\u003c/p\u003e\n\u003ch4\u003eWho was affected?\u003c/h4\u003e\n\u003cp\u003eThis issue affects users with Tailscale SSH enabled who also allow local or remote forwarding in their access policies.\u003c/p\u003e\n\u003ch4\u003eWhat do I need to do?\u003c/h4\u003e\n\u003cp\u003eIf you use Tailscale SSH with local or remote forwarding, upgrade to Tailscale version 1.104.4 or newer.\u003c/p\u003e\n\u003cp\u003eAs a mitigation, you can disable SSH forwarding by setting \u003ccode\u003eTS_SSH_DISABLE_FORWARDING=1\u003c/code\u003e in the \u003ccode\u003etailscaled\u003c/code\u003e service environment and restarting the service. This disables forwarding without disabling all of Tailscale SSH.\u003c/p\u003e"
},
"title": "TS-2026-014",
"title_detail": {
"base": "https://tailscale.com/security-bulletins/index.xml",
"language": null,
"type": "text/plain",
"value": "TS-2026-014"
}
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.