Advisory Database
  • Advisories
  • Dependency Scanning
  1. pypi
  2. ›
  3. pytorch
  4. ›
  5. CVE-2026-24747

CVE-2026-24747: PyTorch Vulnerable to Remote Code Execution via Untrusted Checkpoint Files

January 27, 2026 (updated January 29, 2026)

A vulnerability in PyTorch’s weights_only unpickler allows an attacker to craft a malicious checkpoint file (.pth) that, when loaded with torch.load(..., weights_only=True), can corrupt memory and potentially lead to arbitrary code execution.

References

  • github.com/advisories/GHSA-63cw-57p8-fm3p
  • github.com/pytorch/pytorch
  • github.com/pytorch/pytorch/163122/commit/954dc5183ee9205cbe79876ad05dd2d9ae752139
  • github.com/pytorch/pytorch/issues/163105
  • github.com/pytorch/pytorch/releases/tag/v2.10.0
  • github.com/pytorch/pytorch/security/advisories/GHSA-63cw-57p8-fm3p
  • nvd.nist.gov/vuln/detail/CVE-2026-24747

Code Behaviors & Features

Detect and mitigate CVE-2026-24747 with GitLab Dependency Scanning

Secure your software supply chain by verifying that all open source dependencies used in your projects contain no disclosed vulnerabilities. Learn more about Dependency Scanning →

Affected versions

All versions before 2.10.0

Fixed versions

  • 2.10.0

Solution

Upgrade to version 2.10.0 or above.

Impact 8.8 HIGH

CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H

Learn more about CVSS

Weakness

  • CWE-502: Deserialization of Untrusted Data
  • CWE-94: Improper Control of Generation of Code ('Code Injection')

Source file

pypi/pytorch/CVE-2026-24747.yml

Spotted a mistake? Edit the file on GitLab.

  • Site Repo
  • About GitLab
  • Terms
  • Privacy Statement
  • Contact

Page generated Wed, 04 Feb 2026 00:36:05 +0000.