CVE-2026-1462
highpriority · track · 0EPSS 0.004 · 35 %ileEUUSEPSS · scored 2026-10-27
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.
cvss
Technical severity, 0–10. One row per scoring source.
base score 0–10 · higher = more severe · hover a vector to decode
ssvc
A stakeholder decision — act, attend, or track — from exploitation and impact.
weaknesses
The CWE weakness classes this flaw belongs to.
affected products (4)
Products and version ranges known to be affected.
advisories (2)
Vendor and ecosystem advisories describing this.
- GHSA-4f3f-g24h-fr8mKeras has an untrusted deserialization vulnerability2026-04-13ghsa · osvUS
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.
- PYSEC-2026-2547Keras has an untrusted deserialization vulnerability2026-07-13vendor · osvUS
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.
also known as (6)
Other names and identifiers for this entity.
sources
Which sources assert this record, and when.
references (12)
External citations — patches, advisories, analysis.
- https://github.com/keras-team/keras/commit/b6773d3decaef1b05d8e794458e148cb362f163fcnnvd
- https://github.com/keras-team/keras/releasesPatchcnnvd
- https://huntr.com/bounties/7e78d6f1-6977-4300-b595-e81bdbda331ccnnvd
- https://github.com/keras-team/keras/commit/b6773d3decaef1b05d8e794458e148cb362f163fenisa_euvd
- https://huntr.com/bounties/7e78d6f1-6977-4300-b595-e81bdbda331cenisa_euvd
- https://access.redhat.com/errata/RHSA-2026:24977Third Party Advisorynvd
- https://access.redhat.com/errata/RHSA-2026:37275nvd
- https://access.redhat.com/security/cve/CVE-2026-1462Third Party Advisorynvd
Attackers weaponise vulnerabilities like this against internet-exposed services. See what is exposed on your perimeter with a free passive scan.
Free scan