CVE-2026-115886 · GHSA-2wqp-q3c6-ccj5
MLflow: SSRF on unauthenticated MLflow servers via third-party scorer kwargs in the scorer invoke, register, and GenAI evaluate endpoints
- Risk score
- 52/100
- CVSS
- 6.5 v3.1
- Exploit chance (30 days)
- 0.4%
- Published
- 23 Oct 2026
Open in the live forecast Today's forecast
What to know
- What it is
- MLflow is an open source AI engineering platform for agents, large language models, and machine learning models. From 3.12.0rc0 until 3.17.0, the scorer invoke, register, and GenAI evaluate endpoints, including unauthenticated POST /ajax-api/3.0/mlflow/scorer/invoke, accept serialized third-party scorer data that Scor…
- Affected
- mlflow (pip) >= 3.12.0rc0, < 3.17.0
- Fix
- Upgrade mlflow to 3.17.0 or later.
- Exploitation
- No known exploitation
Story map
At a glance
mlflow: Amber, patch this week. MLflow has a path traversal vulnerability in the FileStore tracking backend, allowing unauthenticated attackers to control filesystem paths and write files with attacker-controlled metadata. Attackers can exploit this to write files outside the intended experiment trace directory.
The flaws
- CVE-2026-115886: MLflow: SSRF on unauthenticated MLflow servers via third-party scorer kwargs in the scorer invoke, register, and GenAI evaluate endpoints (CVSS 6.5 · EPSS 0%)
- CVE-2026-53995: MLflow FileStore trace IDs allow path traversal and arbitrary file writes (CVSS 8.2 · EPSS 1%)
The fix
Upgrade mlflow: 3.17.0 or later on your release line
Who is exposed: {'products': ['MLflow'], 'packages': [{'ecosystem': 'pip', 'name': 'mlflow', 'vulnerable': '<= 3.17.1', 'patched': '3.17.2'}]}
Where it stands
- No known exploitation
Connects to
- Same weakness: CWE-470 (CWE-470). 26 other warnings in 30 days
Why it is Yellow
- CVSS 6.5 medium
- On your watchlist: PyPI
Affected packages
| Package | Ecosystem | Vulnerable | Fixed in |
|---|---|---|---|
| mlflow | pip | >= 3.12.0rc0, < 3.17.0 | 3.17.0 |
References
- https://github.com/advisories/GHSA-2wqp-q3c6-ccj5
- https://github.com/mlflow/mlflow/security/advisories/GHSA-2wqp-q3c6-ccj5
- https://github.com/mlflow/mlflow/pull/25890
- https://github.com/mlflow/mlflow/commit/bac30b4148c4f40045b4d4db07ae60ab3838803f
- https://github.com/mlflow/mlflow/releases/tag/v3.17.0
- https://github.com/advisories/GHSA-2wqp-q3c6-ccj5
Original description
MLflow is an open source AI engineering platform for agents, large language models, and machine learning models. From 3.12.0rc0 until 3.17.0, the scorer invoke, register, and GenAI evaluate endpoints, including unauthenticated POST /ajax-api/3.0/mlflow/scorer/invoke, accept serialized third-party scorer data that Scorer.model_validate_json resolves using a module-path allowlist without constraining the selected class or its constructor arguments. When optional trulens and trulens-providers-litellm packages are installed, an attacker can select mlflow.genai.scorers.trulens, choose a model URI outside MLflow's native provider registry so execution uses the LiteLLM fallback, and supply api_base and api_key values that cause the server to send an outbound request to an attacker-chosen or internal host with attacker-controlled headers and body. The vulnerable deserialization occurs before a tracking-store lookup or job submission. This issue is fixed in version 3.17.0.
Last activity 23 Oct 2026. Patch Forecast updates every hour.