DEV Community

CVE Reports
CVE Reports

Posted on • Originally published at cvereports.com

CVE-2026-69146: CVE-2026-69146: Missing Authorization Bypass in MLflow Basic Authentication Middleware

#security #cve #cybersecurity

CVE-2026-69146: Missing Authorization Bypass in MLflow Basic Authentication Middleware

Vulnerability ID: CVE-2026-69146
CVSS Score: 6.5
Published: 2026-08-17

This technical report details a missing authorization vulnerability (CVE-2026-69146 / GHSA-3p64-6gvh-82v5) affecting the MLflow platform from version 3.13.0 to 3.15.0. When MLflow is configured with the built-in basic-auth plugin, authenticated users can bypass run-level UPDATE authorization checks, enabling unauthorized dataset and model lineage metadata injection.

TL;DR

An authorization bypass vulnerability in MLflow's basic-auth middleware allows authenticated users to inject arbitrary dataset records into other users' runs without UPDATE permissions.

⚠️ Exploit Status: POC

Technical Details

  • CWE ID: CWE-862
  • Attack Vector: Network (AV:N)
  • CVSS: 6.5
  • EPSS Status: Not Populated
  • Impact: High Integrity (I:H)
  • Exploit Status: POC / Test-Suite Verified
  • KEV Status: Not Listed

Affected Systems

  • MLflow Platform Tracking Server
  • MLflow: >= 3.13.0, < 3.15.0 (Fixed in: 3.15.0)

Code Analysis

Commit: 5c34aec

Register LogInputs and LogOutputs in BEFORE_REQUEST_HANDLERS within basic authentication module

Exploit Details

  • MLflow Unit Tests: Unit tests verifying missing validation boundaries on LogInputs and LogOutputs endpoints.

Mitigation Strategies

  • Upgrade to MLflow version 3.15.0 or later to integrate LogInputs and LogOutputs with the tracking server access control middleware.
  • Implement network boundary controls to limit access to the MLflow tracking port.
  • Apply reverse proxy or API gateway rules to filter and deny untrusted external traffic to the log-inputs and log-outputs REST endpoints.

Remediation Steps:

  1. Verify the current MLflow version using package management tools or API server banners.
  2. Upgrade MLflow installations via 'pip install --upgrade mlflow>=3.15.0' or rebuild target containers using the updated image base.
  3. Ensure MLflow tracking servers are configured behind a web application firewall or reverse proxy.
  4. Configure log alerting for HTTP POST requests directed to /api/2.0/mlflow/runs/log-inputs originating from unexpected source networks.

References

Read the full report for CVE-2026-69146 on our website for more details including interactive diagrams and full exploit analysis.

Top comments (0)