PatchSiren

PatchSiren cyber security CVE debrief

CVE-2026-68772 ZenML CVE debrief

CVE-2026-68772 is a remote code execution vulnerability in the CloudpickleMaterializer component of ZenML 0.94.6. Attackers with write access to a shared artifact store can execute arbitrary code by planting a malicious pickle file. The vulnerability is triggered when any user or pipeline materializes the artifact through the unsanitized cloudpickle.load() call in cloudpickle_materializer.py.

Vendor
ZenML
Product
Unknown
CVSS
HIGH 8.5
CISA KEV
Not listed in stored evidence
Original CVE published
2026-08-07
Original CVE updated
2026-09-16
Advisory published
2026-08-07
Advisory updated
2026-09-16

Who should care

Defenders responsible for securing artifact stores and materialization processes in ZenML deployments should assess exposure and prioritize verification of integrity and access controls.

Why it matters

CVE-2026-68772 is a remote code execution vulnerability in ZenML 0.94.6's CloudpickleMaterializer component, allowing attackers with write access to execute arbitrary code. Defenders should prioritize verifying integrity and access controls, and implement input validation and monitoring to prevent exploitation.

  • Verify the integrity of artifact stores to prevent malicious pickle files
  • Restrict write access to trusted users to limit exploitation
  • Implement input validation and sanitization for cloudpickle.load() calls to prevent arbitrary code execution
  • Monitor for suspicious activity in artifact stores and materialization processes to detect potential exploitation

Technical summary

The vulnerability in ZenML 0.94.6's CloudpickleMaterializer component is caused by the unsanitized cloudpickle.load() call in cloudpickle_materializer.py, which allows attackers with write access to a shared artifact store to execute arbitrary code by planting a malicious pickle file. This can be triggered when any user or pipeline materializes the artifact, highlighting the need for defenders to prioritize verifying the integrity of artifact stores, restricting write access to trusted users, and implementing input validation and sanitization for cloudpickle.load() calls.

Defensive priority

Defenders should prioritize verifying the integrity of artifact stores and restricting write access to trusted users.

Recommended defensive actions

  • Verify the integrity of artifact stores and restrict write access to trusted users
  • Implement input validation and sanitization for cloudpickle.load() calls
  • Monitor for suspicious activity in artifact stores and materialization processes
  • Review compensating controls for exposed systems while remediation is scheduled and verified
  • Check relevant monitoring, detection, and logs for exposed assets that need extra review
  • Track exceptions, retest remediated assets, and close the item only after evidence is documented
  • Confirm whether affected product deployments exist in managed environments and assign an owner for follow-up

Evidence notes

The CVE record and NVD entry provide details on the vulnerability in ZenML 0.94.6's CloudpickleMaterializer component, but the scope of affected versions, remediation steps, and potential attack vectors require verification from official sources and defensive testing to confirm exploitability. Defenders should verify the integrity of artifact stores, restrict write access, and monitor for suspicious activity.

Sources and references

Verified primary and authoritative sources

  • CVE-2026-68772 CVE Program record

    Publisher, destination, and source semantics verified

    URL: https://www.cve.org/CVERecord?id=CVE-2026-68772

    CVE Program - Official CVE Program record with source-provided CVE metadata.

  • CVE-2026-68772 NVD vulnerability detail

    Publisher, destination, and source semantics verified

    URL: https://nvd.nist.gov/vuln/detail/CVE-2026-68772

    NIST National Vulnerability Database - Official NIST NVD detail page and source-specific vulnerability assessment.

Supplemental references

Methodology and review provenance

AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.