Advisory

PYSEC-2026-3634

Also known as: CVE-2026-12484, GHSA-v2w2-w228-c444

Summary

Keras: TorchModuleWrapper can deserialize unsafe PyTorch pickle data

Details

A vulnerability in keras-team/keras version 3.15.0 allows unsafe deserialization of attacker-controlled PyTorch pickle data through the public keras.layers.TorchModuleWrapper.from_config method. This method invokes torch.load(..., weights_only=False) without requiring an explicit unsafe opt-in, such as a safe_mode=False parameter. When called outside a SafeModeScope(True) context, the absence of an ambient safe mode state permits unsafe deserialization by default. This issue can lead to arbitrary code execution if untrusted Keras layer configurations are processed using this method. The vulnerability arises because the method does not enforce safe deserialization practices unless explicitly guarded by Keras safe mode.

Severity

High Basis: CVSS v3 score 7.8, vector CVSS:3.0/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H

Dates

Published
2026-08-10 10:43 UTC
Modified
2026-08-10 11:30 UTC

Affected packages and fixes

keras

PyPI · All systems

Affected

  • introduced 0, fixed 3.12.3, introduced 3.13.0, fixed 3.15.0 (ecosystem)

Fixed versions

3.12.3, 3.15.0

Fix

  • Fixed version 3.12.3
    pip install --upgrade keras==3.12.3
  • Fixed version 3.15.0
    pip install --upgrade keras==3.15.0

Fix commands are the export's own generic commands. Ojuju never runs them. Check them against your setup first.

Check your own computer: guard packages check

View this record on OSV.dev

Source

Source: PyPA Advisory Database · CC BY 4.0

Sources & licences