CVE-2025-9906 Affecting kubeflow-pipelines-visualization-server package, versions <2.17.0-r0


Severity

Recommended
0.0
high
0
10

Snyk's Security Team recommends NVD's CVSS assessment. Learn more

Threat Intelligence

EPSS
0.2% (11th percentile)

Do your applications use this vulnerable package?

In a few clicks we can analyze your entire application and see what components are vulnerable in your application, and suggest you quick fixes.

Test your applications
  • Snyk IDSNYK-CHAINGUARDLATEST-KUBEFLOWPIPELINESVISUALIZATIONSERVER-17996903
  • published17 Jul 2026
  • disclosed19 Sept 2025

Introduced: 19 Sep 2025

CVE-2025-9906  (opens in a new tab)

How to fix?

Upgrade Chainguard kubeflow-pipelines-visualization-server to version 2.17.0-r0 or higher.

NVD Description

Note: Versions mentioned in the description apply only to the upstream kubeflow-pipelines-visualization-server package and not the kubeflow-pipelines-visualization-server package as distributed by Chainguard. See How to fix? for Chainguard relevant fixed versions and status.

The Keras Model.load_model method can be exploited to achieve arbitrary code execution, even with safe_mode=True.

One can create a specially crafted .keras model archive that, when loaded via Model.load_model, will trigger arbitrary code to be executed. This is achieved by crafting a special config.json (a file within the .keras archive) that will invoke keras.config.enable_unsafe_deserialization() to disable safe mode. Once safe mode is disable, one can use the Lambda layer feature of keras, which allows arbitrary Python code in the form of pickled code. Both can appear in the same archive. Simply the keras.config.enable_unsafe_deserialization() needs to appear first in the archive and the Lambda with arbitrary code needs to be second.

CVSS Base Scores

version 3.1