Skip to content
Fencer Vulns
high 8.8

CVE-2026-1462 json

Safe Mode Bypass in keras-team/keras

published 2026-04-13 6 mo ago modified 2026-09-10 1 mo ago also GHSA-4f3f-g24h-fr8m , PYSEC-2026-2547 , EUVD-2026-21970

cvss
8.8
version 3.0, CNA
cisa kev
Not listed
not in CISA's catalog
epss
0.4%
EPSS 0.4% chance of exploitation in 30 days, percentile 34%
public exploits
0
none known

Summary

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.

Patch / fix

  • pypi keras 3.13.2
  • keras-team keras-team/keras 3.13.2
  • Red Hat Red Hat OpenShift AI 2.25 1780394782
  • Red Hat Red Hat OpenShift AI 3.3 1782471730, 1782471734, 1782471835, 1782471879

Vendor advisories

What the vendors themselves say about their products: affected, fixed or not affected, and what to do.

red hat product security vex CVE-2026-1462 updated cvss 7.8

keras: Keras: Arbitrary Code Execution Vulnerability Bypassing Safe Mode

  • Red Hat OpenShift AI (RHOAI) rhoai/odh-modelmesh-runtime-adapter-rhel8
    affected: workaround, no fix available
  • Red Hat OpenShift AI 2.25 rhoai/odh-modelmesh-runtime-adapter-rhel9 1780394782
    fixed: vendor fix RHSA-2026:24977
  • Red Hat OpenShift AI 3.3 rhoai/odh-pipeline-runtime-tensorflow-cuda-py312-rhel9 1782471734, rhoai/odh-pipeline-runtime-tensorflow-rocm-py312-rhel9 1782471879, rhoai/odh-workbench-jupyter-tensorflow-cuda-py312-rhel9 1782471730, rhoai/odh-workbench-jupyter-tensorflow-rocm-py312-rhel9 1782471835
    fixed: vendor fix RHSA-2026:37275
  • Red Hat OpenShift AI (RHOAI) rhoai/odh-kserve-agent-rhel9, rhoai/odh-kserve-router-rhel9, rhoai/odh-kserve-controller-rhel9, rhoai/odh-kserve-storage-initializer-rhel9
    not affected: vulnerable code not in execute path
  • Red Hat OpenShift AI 2.25 rhoai/odh-rhel9-operator 1780513840, rhoai/odh-vllm-cpu-rhel9 1780069252, rhoai/odh-dashboard-rhel9 1780467029, rhoai/odh-modelmesh-rhel9 1780069506, rhoai/odh-operator-bundle 1780514621, rhoai/odh-vllm-cuda-rhel9 1780069320, +70 more Red Hat OpenShift AI 3.3 rhoai/odh-cli-rhel9 1782471642, rhoai/rhai-cli-rhel9 1782471642, rhoai/odh-mlflow-rhel9 1782917849, rhoai/odh-trainer-rhel9 1782471656, rhoai/odh-maas-api-rhel9 1782996644, rhoai/odh-mlserver-rhel9 1782887848, +73 more
    not affected: vulnerable code not present

Red Hat CSAF VEX © Red Hat, Inc., from https://security.access.redhat.com/data/csaf/v2/vex/; changed: reduced to product statuses and remediations. CC BY 4.0 · csaf

CISA SSVC

Exploitation
Proof of concept is public
Automatable
no
Technical impact
total
Assessed
6 mo ago

Exploit prediction

0.40% chance of exploitation in the next 30 days, higher than 34% of all scored CVEs (FIRST EPSS, 2026-11-08).

Exploits and checks

  • huntr.com https://huntr.com/bounties/7e78d6f1-6977-4300-b595-e81bdbda331c referenced PoC, tagged "exploit" by a source, unverified

CVSS scores

assigned by version score vector
CNA 3.0 8.8 CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
CISA ADP 3.1 7.8 CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
CNA 3.1 7.8 CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
GitHub 3.0 8.8 CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
OSV 3.0 8.8 CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H

Affected packages

ecosystem package affected fixed in
pypi keras
<3.13.2
3.13.2

Affected products

  • keras-team keras-team/keras
  • Red Hat Red Hat OpenShift AI 2.25
  • Red Hat Red Hat OpenShift AI 3.3
  • Red Hat Red Hat OpenShift AI (RHOAI)

Classification

Weakness
  • CWE-502 Deserialization of Untrusted Data

References

Sources

Fencer Vulns keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.