CVE-2026-116692 mlflow CVE debrief

Who should care

Defenders responsible for MLflow servers, especially those with public access or unauthenticated access, should assess exposure and prioritize verification and potential remediation.

Why it matters

CVE-2026-116692 allows unauthenticated remote code execution on MLflow servers via third-party scorer deserialization. Defenders should prioritize verifying exposure, especially for public or unauthenticated MLflow servers, and assess the feasibility of exploitation in their environment. The vulnerability was introduced in MLflow 3.12.0 and fixed in version 3.17.0.

• Unauthenticated remote code execution with the privileges of the MLflow server process
• Potential for arbitrary Python code execution
• Possible exploitation through MLflow's built-in endpoints
• Requires verification of exposure and environment-specific feasibility

Technical summary

MLflow 3.12.0 introduced registration and deserialization support for third-party scorers. An HTTP caller can combine this behavior with MLflow's built-in experiment, run, and artifact-upload endpoints to achieve unauthenticated arbitrary Python code execution with the privileges of the MLflow server process on the default no-auth server using the official MLflow v3.14.0 image. This execution occurs because the server accepts a request-controlled Python module path beneath one of four allowed package prefixes, imports that module, obtains a request-controlled class, and instantiates it synchronously.

Defensive priority

Defenders should prioritize verifying exposure of MLflow servers, especially those with public access or unauthenticated access, and assess the feasibility of exploitation given their environment.

Recommended defensive actions

• Verify MLflow server exposure and authentication settings
• Assess the feasibility of exploitation in the environment
• Update to version 3.17.0 or later if possible
• Monitor for suspicious activity related to MLflow scorer deserialization
• 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

Evidence notes

The CVE record and source item provide details on the vulnerability in MLflow, including the affected versions and potential impact. However, some details, such as specific exploitation scenarios or victim information, are not provided.

Sources and 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.