CVE-2026-8147 in MLflow
In MLflow versions prior to 3.14.0, when running with authentication enabled, the trace API endpoints lack proper authorization validators. This allows any authenticated user to bypass experiment-level authorization controls on all trace operations, including reading, deleting, and modifying traces on experiments they do not have permission to access. The issue arises from the `_before_request` handler, which does not register authorization validators for trace endpoints, resulting in requests proceeding without validation. This vulnerability can expose sensitive data, destroy audit logs, and allow unauthorized modifications.
Quick answer
LF Projects MLflow should be reviewed and updated if it matches the affected versions. The recommended fix is to apply the vendor-supported patched version or the mitigation steps below, then retest the public website with Fixnx.
Who is affected
Affected versions
- MLflow before 3.14.0 when authentication is enabled
Fixed versions
- MLflow 3.14.0 or later
How to fix it
CVE-2026-8147 affects MLflow. The issue means any authenticated user may bypass trace API authorization and read, change, or delete traces. Treat it as urgent if the system is reachable by users, browsers, admins, or the network. Update to MLflow 3.14.0 or later, or follow the vendor advisory if your branch needs a supported backport.
- Inventory every MLflow install, device, browser fleet, container, or package in use.
- Compare installed versions with the affected ranges in the vendor advisory.
- Update to MLflow 3.14.0 or later on supported systems.
- If you cannot update today, limit access to trusted users and trusted networks only.
- Review logs for unusual access, crashes, privilege changes, commands, or failed requests.
- Rotate secrets, tokens, and service accounts if abuse or data access is suspected.
- Document the patch and keep the advisory link with the change record.
Scan now. Google sign-in is only needed to unlock fix guidance.
Verify the fix
- Confirm the installed version is MLflow 3.14.0 or later.
- Confirm exposed admin, API, browser, or device access is limited to trusted users.
- Retest the affected workflow after the update or mitigation.
- Review logs again after patching for repeated suspicious activity.
- Run a new Fixnx scan where the affected service is part of the public site.
Related categories
Trusted references
FAQ
What is affected by CVE-2026-8147?
LF Projects MLflow versions listed as affected should be reviewed: MLflow before 3.14.0 when authentication is enabled.
What should I fix first?
Start with internet-facing sites, admin panels, login flows, plugins, themes, modules, packages, and systems that process user-controlled input or sensitive data.
How do I confirm the fix worked?
Apply the patched version or mitigation, clear caches where relevant, retest the affected workflow, and run a new Fixnx scan to verify public website exposure signals.
How are Fixnx security risk categories chosen?
Fixnx keeps one canonical risk page and assigns only broad, relevant categories such as ecosystem, technology area, or vulnerability class.