CVE-2026-61682

9.9

kcp-dev · kcp

The kcp front-proxy fails to sanitize identity headers, allowing authenticated tenants to perform cross-workspace impersonation and gain elevated privileges.

Executive summary

A critical authentication bypass vulnerability in the kcp control plane allows authenticated users to escalate privileges to system administrator levels, posing a severe risk to cluster security.

Vulnerability

This flaw stems from the failure of the kcp front-proxy to strip inbound identity headers like X-Remote-Group before forwarding requests to shards. An authenticated tenant can inject headers to spoof identity assertions, effectively bypassing authorization checks and gaining control over secrets, RBAC data, and cluster resources.

Business impact

The ability for a standard tenant to impersonate privileged users results in total compromise of the Kubernetes-like control plane. Given the CVSS score of 9.9, this vulnerability represents an existential threat to multi-tenant environments, as attackers can exfiltrate sensitive secrets, modify infrastructure configurations, or delete critical business logic.

Remediation

Immediate Action: Upgrade the kcp deployment to version 0.31.4 or 0.32.2 immediately to implement the required header sanitization logic.

Proactive Monitoring: Audit Kubernetes and kcp access logs for suspicious identity header injection patterns or unauthorized cross-workspace resource access.

Compensating Controls: Ensure strict network segmentation for the kcp front-proxy and utilize mTLS to verify identity assertions between components if the patch cannot be applied immediately.

Exploitation status

Public Exploit Available: Unknown

Analyst recommendation

Organizations utilizing kcp must prioritize this update as a critical security task. The vulnerability allows for full administrative bypass, meaning any delay in patching significantly increases the risk of a total control plane takeover by malicious tenants.

History

  1. Disclosed CVE record published
  2. Collected by CVE Brief via github
  3. Analyst report written
  4. Published in the daily brief critical section

Sources