mlflow · MLflow · published 2026-10-27
CVE-2026-116692MLflow: Remote code execution on unauthenticated MLflow servers via third-party scorer deserialization importing caller-placed artifact modules
This month Public proof-of-concept exists (SSVC exploitation: poc).
CVE-2026-116692 has a public proof-of-concept exploit, but no confirmed exploitation in the wild. EPSS estimates a 2.4% chance of exploitation in the next 30 days, higher than 84% of all CVEs. This MLflow vulnerability: CVSS 9.8 (v3.1), attacks can be automated, a successful attack gives total control of the component. Fixed upstream in MLflow 3.17.0. Patch in your normal monthly cycle.
Jump to: Remediation Exploit intelAffected versionsDescription
Remediation#
Fixed upstream in MLflow 3.17.0.
Patch events
AI toolingMLflow
MLflow security release: 3.17.0
Upgrade to
3.17.0Why: Public proof-of-concept exists (SSVC exploitation: poc).
Exploit intelligence#
| Signal | Value | Source |
|---|---|---|
| Known exploited | No | CISA KEV |
| Ransomware use | — | CISA KEV |
| EPSS (30-day exploitation probability) | 2.4% (84th pct) | FIRST EPSS v2026.06.15 |
| CVSS base score | 9.8 (v3.1)CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H | CNA / CISA ADP |
| SSVC exploitation | poc | CISA Vulnrichment |
| SSVC automatable | Yes | CISA Vulnrichment |
| SSVC technical impact | total | CISA Vulnrichment |
Affected versions (from the CVE record)#
- mlflow mlflow
>= 3.12.0rc0, < 3.17.0(affected)
Description (from the CVE record)#
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, third-party scorer deserialization in mlflow/genai/scorers/base.py accepts request-selected descendants of THIRD_PARTY_SCORER_ALLOWED_MODULES rather than requiring an exact allowed module. A caller can use MLflow artifact endpoints to place a Python module beneath a writable allowed package and submit its module and class through /ajax-api/3.0/mlflow/scorer/invoke. Scorer.model_validate_json() imports the module and constructs the class before trace processing. Authentication-disabled deployments permit this code execution without credentials, while basic-auth deployments require a principal capable of creating an experiment and run and uploading artifacts. In either configuration, the allowed scorer package must be writable by the server process. This issue is fixed in version 3.17.0.
Quoted from the CVE Program record.