Kyverno Cross-Namespace Privilege Escalation via Policy apiCall in github.com/kyverno/kyverno

The core of the vulnerability lies within the apiCall policy feature in Kyverno. An attacker can craft a malicious policy that, when applied, grants them elevated permissions beyond their intended scope. This can manifest as the ability to read, modify, or delete resources in namespaces they should not have access to. The impact is significant, potentially allowing an attacker to compromise the entire Kubernetes cluster if they can successfully exploit this vulnerability and gain control of a privileged account. This is particularly concerning in multi-tenant environments where namespaces isolate different applications or teams.

Organizations heavily reliant on Kyverno for Kubernetes policy enforcement are at significant risk. This includes those using Kyverno to enforce strict security policies, manage access control, or automate deployments. Shared Kubernetes environments and those with complex policy configurations are particularly vulnerable.

• linux / server:

journalctl -u kyverno -f | grep -i "apiCall"

• go / supply-chain: Inspect Kyverno policy files for instances of apiCall with potentially insecure configurations. Look for policies that allow unrestricted access to Kubernetes API resources. • generic web: Monitor Kubernetes API audit logs for unusual patterns of API calls originating from Kyverno pods, particularly those involving resource modifications or privilege escalations.

1. 2026-02-02Disclosure

   disclosure

1. Reserviert2026-01-05
2. Veröffentlicht2026-02-02
3. Geändert2026-02-04
4. EPSS aktualisiert2026-03-23

The primary mitigation is to immediately upgrade Kyverno to version 1.15.3 or later, which includes the fix for this vulnerability. If upgrading is not immediately feasible, consider temporarily restricting the use of apiCall policies, especially those with broad scopes. Review existing apiCall policies to identify any that might be susceptible to exploitation. Monitor Kyverno logs for any unusual activity related to policy evaluation or resource access. While a WAF cannot directly prevent this vulnerability, it can help detect and block malicious requests attempting to exploit it.