Kurz gesagt
- CVE-2026-1462 ist eine Schwachstelle (HOCH (CVSS 7.8)) in keras, keras (keras), openshift ai (redhat).
- Wahrscheinlichkeit einer Ausnutzung in den nächsten 30 Tagen (EPSS): 0.40%, höher als bei 34.7% aller Schwachstellen.
- Abhilfe: auf 3.13.2, 2.25.7 oder neuer aktualisieren.
| Package | Affected versions | Patched versions |
|---|
| keras PyPI | < 3.13.2
| 3.13.2
|
| keras | = 3.13.0
| No fix yet |
| openshift ai | ≥ 2.25, < 2.25.7
| 2.25.7
|
Description
A vulnerability in the TFSMLayer class of the keras package, version 3.13.0, allows attacker-controlled TensorFlow SavedModels to be loaded during deserialization of .keras models, even when safe_mode=True. This bypasses the security guarantees of safe_mode and enables arbitrary attacker-controlled code execution during model inference under the victim's privileges. The issue arises due to the unconditional loading of external SavedModels, serialization of attacker-controlled file paths, and the lack of validation in the from_config() method.
NVD description
A vulnerability in the `TFSMLayer` class of the `keras` package, version 3.13.0, allows attacker-controlled TensorFlow SavedModels to be loaded during deserialization of `.keras` models, even when `safe_mode=True`. This bypasses the security guarantees of `safe_mode` and enables arbitrary attacker-controlled code execution during model inference under the victim's privileges. The issue arises due to the unconditional loading of external SavedModels, serialization of attacker-controlled file paths, and the lack of validation in the `from_config()` method.
Running one of these products? Scan your site to see whether it exposes a vulnerable version.
Scan my site free