CVE-2026-2033High· 8.1▾ Abyssal0dayMLflow Tracking Server Artifact Handler Directory Traversal Remote Code Execution Vulnerability
▾ Abyssal zone — Critical with a public exploit or in-the-wild use
impact 44.6 · likelihood 0.3 · exploitation 25
Need a working PoC? Pro members can cast a request and our team develops one — it lands right here.
Exploit-prediction probability, daily snapshots since Jul 13.
Disclosure to exploitation, from the record and what we observed since indexing it.
Disclosed via OSV
1.7%
Last analysed / modified upstream
1.7% → 1.7%
MLflow Tracking Server Artifact Handler Directory Traversal Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of MLflow Tracking Server. Authentication is not required to exploit this vulnerability.
The specific flaw exists within the handling of artifact file paths. The issue results from the lack of proper validation of a user-supplied path prior to using it in file operations. An attacker can leverage this vulnerability to execute code in the context of the service account.
mlflow < 3.8.0rc0Upgrade to a patched release:
mlflow 3.8.0rc0Connected by shared product, vendor, weakness, or advisory.
CVE-2025-14287High· 7.5MLflow has a command injection in mlflow/sagemaker/__init__.py
CVE-2025-15036Critical· 9.6MLFlow path traversal vulnerability
CVE-2026-33866Medium· 4.3MLflow is vulnerable to an authorization bypass affecting the AJAX endpoint
CVE-2025-15379Critical· 10.0A command injection vulnerability exists in MLflow's model serving container initialization code, specifically in the `_install_model_dependencies_to_env()` function
CVE-2026-8147High· 8.1MLflow: trace API endpoints lack proper authorization validators
CVE-2026-71211High· 7.1MLflow's AI Gateway accepts an auth_config.api_base value when creating a gateway secret (mlflow/server/handlers.py, _create_gateway_secret) with no validation of scheme, host, or IP range; the value is stored verbatim