CVE-2026-1462
A vulnerability in the `TFSMLayer` class of the `keras` package, version 3.13.0, allows attacker-controlled TensorFlow SavedModels to be loa
CVSS
7.8
Alto
EPSS
0.4%
p35
KEV
—
Exploit Today
10
0-100
Publicado: 13 abr 2026 · Última mod.: 15 jul 2026 · CWE-502
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.
- github.comhttps://github.com/keras-team/keras/commit/b6773d3decaef1b05d8e794458e148cb362f163f
- huntr.comhttps://huntr.com/bounties/7e78d6f1-6977-4300-b595-e81bdbda331c
- access.redhat.comhttps://access.redhat.com/errata/RHSA-2026:24977
- access.redhat.comhttps://access.redhat.com/errata/RHSA-2026:37275
- access.redhat.comhttps://access.redhat.com/security/cve/CVE-2026-1462
- bugzilla.redhat.comhttps://bugzilla.redhat.com/show_bug.cgi?id=2457856
- security.access.redhat.comhttps://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-1462.json