Skip to main content
← All vulnerabilities
CVE-2026-92574

Cri-o: cri-o checkpoint restore bypasses destination security context

Affects CRI-O 1.34.0 to before 1.34.14, 1.35.0 to before 1.35.9, and 1.36.0 to before 1.36.6 where checkpoint restore is available, which is on by default. Fixed in 1.34.14, 1.35.9, and 1.36.6.

Published
Last modified
published
Not listed in CISA KEV
Specific setups only
Fix
Fixed in 1.34.14, 1.35.9, 1.36.6
None on Exploit-DB, Metasploit or GitHub
  • Exploitation: none
  • Automatable: no
  • Technical impact: total

Summary

A vulnerability in CRI-O checkpoint restore allows a user who can create a pod from a malicious checkpointed container to bypass the destination Kubernetes security context. The restored process may retain credentials, Linux capabilities, no_new_privs, and seccomp state from the checkpoint instead of enforcing the destination configuration. This can allow execution with elevated privileges across the container security boundary. Affected upstream supported versions are CRI-O 1.34 and later. Downstream Red Hat products are affected from OCP 4.17 onward. Fixes have been applied to supported branches but are not yet released. Exploitation requires permission to create a pod from a malicious checkpoint image and checkpoint restore functionality to be available.

Timeline

  1. CVE published
  2. CISA SSVC assessment

    Exploitation: none · Automatable: no · Technical impact: total

  3. Record last updated

Affected versions

  • cri-o

    From 1.34.0 · fixed in 1.34.14; From 1.35.0 · fixed in 1.35.9; From 1.36.0 · fixed in 1.36.6

  • openshift-sandboxed-containers/osc-monitor-rhel9

    Version range not reported

  • container-tools:rhel8/conmon

    Version range not reported

  • container-tools:rhel8/podman

    Version range not reported

  • conmon

    Version range not reported

  • conmon

    Version range not reported

  • cri-o

    Version range not reported

  • kata-containers

    Version range not reported

  • openshift4/cnf-tests-rhel8

    Version range not reported

  • openshift4/microshift-bootc-rhel10

    Version range not reported

  • openshift4/microshift-bootc-rhel9

    Version range not reported

  • openshift4/ztp-site-generate-rhel8

    Version range not reported

  • openshift/ose-rhel-coreos-8

    Version range not reported

  • openshift/ose-rhel-coreos-9

    Version range not reported

  • ose-crio-credential-provider

    Version range not reported

  • cri-o

    Version range not reported

  • kata-containers

    Version range not reported

  • openshift5/microshift-bootc-rhel10

    Version range not reported

  • openshift5/microshift-bootc-rhel9

    Version range not reported

  • openshift/ose-rhel-coreos-9

    Version range not reported

  • ose-crio-credential-provider

    Version range not reported

References 2

access.redhat.com

bugzilla.redhat.com

FAQ

Questions about CVE-2026-92574

What does this mean?

A pod created from a malicious checkpoint image can run its processes as root with full capabilities and no seccomp filtering. This bypasses the security context the orchestrator requested for the destination container.Source: access.redhat.com

Am I affected, and how do I check?

Upstream CRI-O 1.34 and later are affected, up to the fixed releases 1.34.14, 1.35.9, and 1.36.6 on their branches. Red Hat OpenShift Container Platform is affected from 4.17 onward, and checkpoint restore from remote registries is enabled by default in affected versions. Check your CRI-O version and whether enable_criu_support is set in crio.conf, or whether the ContainerCheckpoint feature gate is enabled in OpenShift.Source: access.redhat.com

What should I do about it?

Update CRI-O to 1.36.6 on the 1.36.x branch, 1.35.9 on the 1.35.x branch, or 1.34.14 on the 1.34.x branch. If you cannot update yet, disable checkpoint restore as described in the no-patch answer.Source: access.redhat.com

Is it being exploited?

It is not in CISA's Known Exploited Vulnerabilities catalog, and CISA's SSVC assessment does not mark exploitation as active. That does not mean it has not been attacked: vendors and researchers can report attacks before CISA lists them. No public exploit code is tracked.Source: cve.now feed data