CVE-2025-9905: Arbitary Code execution in Keras load_model()

Severity: high (CVSS 7.3 v3.1)
Exploited in the wild: not listed in CISA KEV
EPSS: 0.2% probability of exploitation in the next 30 days
Public exploit or detection template: yes
Published: 2025-09-19 · updated 2026-10-19
Weakness: CWE-913

Description

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 .h5/.hdf5 model archive that, when loaded via Model.load_model, will trigger arbitrary code to be executed. This is achieved by crafting a special .h5 archive file that uses the Lambda layer feature of keras which allows arbitrary Python code in the form of pickled code. The vulnerability comes from the fact that the safe_mode=True option is not honored when reading .h5 archives. Note that the .h5/.hdf5 format is a legacy format supported by Keras 3 for backwards compatibility.

Affected products

keras-team/keras: from 3.0.0 before 3.11.3 — fixed in 3.11.3

How to fix

Upgrade to a fixed version: keras-team/keras 3.11.3. Check the vendor advisory for mitigations.

References

Get alerted when CVE-2025-9905 or the next vulnerability affects your stack