GCVE Workshop - 22 September 2026 (14:00-18:00), Luxembourg Before The Vulnopticon Conference - Registration

PYSEC-2026-1856

Vulnerability from pysec - Published: 2026-07-07 16:03 - Updated: 2026-07-07 17:25
VLAI
Details

Summary

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.

Vulnerability Details

The weights_only=True unpickler failed to properly validate pickle opcodes and storage metadata, allowing:

  1. Heap memory corruption via SETITEM/SETITEMS opcodes applied to non-dictionary types
  2. Storage size mismatch between declared element count and actual data in the archive

Impact

An attacker who can convince a user to load a malicious checkpoint file may achieve arbitrary code execution in the context of the victim's process.

Credit

Ji'an Zhou

Severity
8.8 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
Impacted products
Name purl
pytorch pkg:pypi/pytorch
Aliases
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "pytorch",
        "purl": "pkg:pypi/pytorch"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "2.10.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ],
      "versions": [
        "0.1.2",
        "1.0.2"
      ]
    }
  ],
  "aliases": [
    "CVE-2026-24747",
    "GHSA-63cw-57p8-fm3p"
  ],
  "details": "### Summary\n\nA vulnerability in PyTorch\u0027s `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.\n\n### Vulnerability Details\n\nThe `weights_only=True` unpickler failed to properly validate pickle opcodes and storage metadata, allowing:\n\n1. **Heap memory corruption** via `SETITEM`/`SETITEMS` opcodes applied to non-dictionary types\n2. **Storage size mismatch** between declared element count and actual data in the archive\n\n### Impact\n\nAn attacker who can convince a user to load a malicious checkpoint file may achieve arbitrary code execution in the context of the victim\u0027s process.\n\n\n# Credit\nJi\u0027an Zhou",
  "id": "PYSEC-2026-1856",
  "modified": "2026-07-07T17:25:15.153263Z",
  "published": "2026-07-07T16:03:20.837738Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/pytorch/pytorch/security/advisories/GHSA-63cw-57p8-fm3p"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-24747"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pytorch/pytorch/issues/163105"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/pytorch/pytorch"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pytorch/pytorch/163122/commit/954dc5183ee9205cbe79876ad05dd2d9ae752139"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pytorch/pytorch/releases/tag/v2.10.0"
    },
    {
      "type": "PACKAGE",
      "url": "https://pypi.org/project/pytorch"
    },
    {
      "type": "ADVISORY",
      "url": "https://github.com/advisories/GHSA-63cw-57p8-fm3p"
    }
  ],
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "PyTorch Vulnerable to Remote Code Execution via Untrusted Checkpoint Files"
}



Log in or create an account to share your comment.




Tags
Taxonomy of the tags.





Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.

Sightings

Author Source Type Date Other

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.


Detection rules are retrieved from Rulezet.


Related by attack behaviour

Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.