Vulnerability-Lookup

CVE-2026-71281 (GCVE-0-2026-71281)

Vulnerability from cvelistv5 – Published: 2026-08-05 12:26 – Updated: 2026-08-10 11:47

Title

peft Unsafe Deserialization via torch.load() Without weights_only in LoRA-GA and CorDA Modules

Summary

Severity

8.8 (High)


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

SSVC

Exploitation: none Automatable: no Technical Impact: total

CISA Coordinator · CISA-ADP (v2.0.3)

Decision recorded 2026-08-05 15:51 UTC

CWE

CWE-502

Assigner

TuranSec redhat root

References

2 references

URL	Tags
https://github.com/huggingface/peft	third-party-advisory
https://github.com/huggingface/peft/blob/main/src…	third-party-advisory

Impacted products

1 product

Vendor	Product	Version	CPE status
huggingface	peft	Affected: 0 , ≤ 0.19.1 (semver)	guessed

Date Public

2026-08-05 12:26

Credits

Saidakbarxon Maxsudxonov

Show details on NVD website

JSON

To clipboard

{
  "containers": {
    "adp": [
      {
        "metrics": [
          {
            "other": {
              "content": {
                "id": "CVE-2026-71281",
                "options": [
                  {
                    "Exploitation": "none"
                  },
                  {
                    "Automatable": "no"
                  },
                  {
                    "Technical Impact": "total"
                  }
                ],
                "role": "CISA Coordinator",
                "timestamp": "2026-08-05T15:51:18.837851Z",
                "version": "2.0.3"
              },
              "type": "ssvc"
            }
          }
        ],
        "providerMetadata": {
          "dateUpdated": "2026-08-05T15:51:29.940Z",
          "orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
          "shortName": "CISA-ADP"
        },
        "title": "CISA ADP Vulnrichment"
      }
    ],
    "cna": {
      "affected": [
        {
          "defaultStatus": "unaffected",
          "product": "peft",
          "programFiles": [
            "src/peft/tuners/lora/corda.py",
            "src/peft/tuners/lora/loraga.py"
          ],
          "vendor": "huggingface",
          "versions": [
            {
              "lessThanOrEqual": "0.19.1",
              "status": "affected",
              "version": "0",
              "versionType": "semver"
            }
          ]
        }
      ],
      "credits": [
        {
          "lang": "en",
          "type": "finder",
          "value": "Saidakbarxon Maxsudxonov"
        }
      ],
      "datePublic": "2026-08-05T12:26:27.376Z",
      "descriptions": [
        {
          "lang": "en",
          "supportingMedia": [
            {
              "base64": false,
              "type": "text/html",
              "value": "\u003cp\u003eHugging Face peft\u0027s LoRA-GA and CorDA initialization modules (src/peft/tuners/lora/corda.py lines ~102 and ~163, and src/peft/tuners/lora/loraga.py line ~101) call torch.load on config-specified cache/covariance files without weights_only=True, bypassing peft\u0027s own safe-loading wrapper used elsewhere in the codebase.\u003c/p\u003e"
            }
          ],
          "value": "Hugging Face peft\u0027s LoRA-GA and CorDA initialization modules (src/peft/tuners/lora/corda.py lines ~102 and ~163, and src/peft/tuners/lora/loraga.py line ~101) call torch.load on config-specified cache/covariance files without weights_only=True, bypassing peft\u0027s own safe-loading wrapper used elsewhere in the codebase."
        }
      ],
      "metrics": [
        {
          "cvssV3_1": {
            "attackComplexity": "LOW",
            "attackVector": "NETWORK",
            "availabilityImpact": "HIGH",
            "baseScore": 8.8,
            "baseSeverity": "HIGH",
            "confidentialityImpact": "HIGH",
            "integrityImpact": "HIGH",
            "privilegesRequired": "NONE",
            "scope": "UNCHANGED",
            "userInteraction": "REQUIRED",
            "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
            "version": "3.1"
          }
        }
      ],
      "problemTypes": [
        {
          "descriptions": [
            {
              "cweId": "CWE-502",
              "description": "CWE-502",
              "lang": "en",
              "type": "CWE"
            }
          ]
        }
      ],
      "providerMetadata": {
        "dateUpdated": "2026-08-10T11:47:04.582Z",
        "orgId": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
        "shortName": "TuranSec"
      },
      "references": [
        {
          "tags": [
            "third-party-advisory"
          ],
          "url": "https://github.com/huggingface/peft"
        },
        {
          "tags": [
            "third-party-advisory"
          ],
          "url": "https://github.com/huggingface/peft/blob/main/src/peft/tuners/lora/corda.py"
        }
      ],
      "source": {
        "discovery": "EXTERNAL"
      },
      "title": "peft Unsafe Deserialization via torch.load() Without weights_only in LoRA-GA and CorDA Modules"
    }
  },
  "cveMetadata": {
    "assignerOrgId": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
    "assignerShortName": "TuranSec",
    "cveId": "CVE-2026-71281",
    "datePublished": "2026-08-05T12:26:27.376Z",
    "dateReserved": "2026-08-05T12:23:34.968Z",
    "dateUpdated": "2026-08-10T11:47:04.582Z",
    "state": "PUBLISHED"
  },
  "dataType": "CVE_RECORD",
  "dataVersion": "5.2",
  "vulnerability-lookup:meta": {
    "epss": {
      "cve": "CVE-2026-71281",
      "date": "2026-09-18",
      "epss": "0.00272",
      "percentile": "0.19783"
    },
    "nvd": {
      "cve": {
        "affected": [
          {
            "affectedData": [
              {
                "defaultStatus": "unaffected",
                "product": "peft",
                "programFiles": [
                  "src/peft/tuners/lora/corda.py",
                  "src/peft/tuners/lora/loraga.py"
                ],
                "vendor": "huggingface",
                "versions": [
                  {
                    "lessThanOrEqual": "0.19.1",
                    "status": "affected",
                    "version": "0",
                    "versionType": "semver"
                  }
                ]
              }
            ],
            "source": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c"
          }
        ],
        "cveTags": [],
        "descriptions": [
          {
            "lang": "en",
            "value": "Hugging Face peft\u0027s LoRA-GA and CorDA initialization modules (src/peft/tuners/lora/corda.py lines ~102 and ~163, and src/peft/tuners/lora/loraga.py line ~101) call torch.load on config-specified cache/covariance files without weights_only=True, bypassing peft\u0027s own safe-loading wrapper used elsewhere in the codebase."
          }
        ],
        "id": "CVE-2026-71281",
        "lastModified": "2026-08-26T17:13:24.800",
        "metrics": {
          "cvssMetricV31": [
            {
              "cvssData": {
                "attackComplexity": "LOW",
                "attackVector": "NETWORK",
                "availabilityImpact": "HIGH",
                "baseScore": 8.8,
                "baseSeverity": "HIGH",
                "confidentialityImpact": "HIGH",
                "integrityImpact": "HIGH",
                "privilegesRequired": "NONE",
                "scope": "UNCHANGED",
                "userInteraction": "REQUIRED",
                "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
                "version": "3.1"
              },
              "exploitabilityScore": 2.8,
              "impactScore": 5.9,
              "source": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
              "type": "Secondary"
            }
          ],
          "ssvcV203": [
            {
              "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
              "ssvcData": {
                "id": "CVE-2026-71281",
                "options": [
                  {
                    "exploitation": "none"
                  },
                  {
                    "automatable": "no"
                  },
                  {
                    "technicalImpact": "total"
                  }
                ],
                "role": "CISA Coordinator",
                "timestamp": "2026-08-05T15:51:18.837851Z",
                "version": "2.0.3"
              }
            }
          ]
        },
        "published": "2026-08-05T13:24:52.723",
        "references": [
          {
            "source": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
            "url": "https://github.com/huggingface/peft"
          },
          {
            "source": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
            "url": "https://github.com/huggingface/peft/blob/main/src/peft/tuners/lora/corda.py"
          }
        ],
        "sourceIdentifier": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
        "vulnStatus": "Deferred",
        "weaknesses": [
          {
            "description": [
              {
                "lang": "en",
                "value": "CWE-502"
              }
            ],
            "source": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
            "type": "Secondary"
          }
        ]
      }
    },
    "redhat_vex": {
      "aggregate_severity": "Important",
      "current_release_date": "2026-09-08T21:57:42+00:00",
      "cve": "CVE-2026-71281",
      "id": "CVE-2026-71281",
      "initial_release_date": "2026-08-05T12:26:27.376000+00:00",
      "product_status:fixed": "4",
      "product_status:known_affected": "24",
      "product_status:known_not_affected": "206",
      "source": "Red Hat CSAF VEX",
      "status": "final",
      "title": "peft: peft: Arbitrary Code Execution via Unsafe Deserialization in LoRA-GA and CorDA Modules",
      "url": "https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-71281.json",
      "version": "3"
    },
    "vulnrichment": {
      "containers": {
        "adp": [
          {
            "metrics": [
              {
                "other": {
                  "content": {
                    "id": "CVE-2026-71281",
                    "options": [
                      {
                        "Exploitation": "none"
                      },
                      {
                        "Automatable": "no"
                      },
                      {
                        "Technical Impact": "total"
                      }
                    ],
                    "role": "CISA Coordinator",
                    "timestamp": "2026-08-05T15:51:18.837851Z",
                    "version": "2.0.3"
                  },
                  "type": "ssvc"
                }
              }
            ],
            "providerMetadata": {
              "dateUpdated": "2026-08-05T15:51:22.177Z",
              "orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
              "shortName": "CISA-ADP"
            },
            "title": "CISA ADP Vulnrichment"
          }
        ],
        "cna": {
          "affected": [
            {
              "defaultStatus": "unaffected",
              "product": "peft",
              "programFiles": [
                "src/peft/tuners/lora/corda.py",
                "src/peft/tuners/lora/loraga.py"
              ],
              "vendor": "huggingface",
              "versions": [
                {
                  "lessThanOrEqual": "0.19.1",
                  "status": "affected",
                  "version": "0",
                  "versionType": "semver"
                }
              ]
            }
          ],
          "credits": [
            {
              "lang": "en",
              "type": "finder",
              "value": "Saidakbarxon Maxsudxonov"
            }
          ],
          "datePublic": "2026-08-05T12:26:27.376Z",
          "descriptions": [
            {
              "lang": "en",
              "supportingMedia": [
                {
                  "base64": false,
                  "type": "text/html",
                  "value": "\u003cp\u003eHugging Face peft\u0027s LoRA-GA and CorDA initialization modules (src/peft/tuners/lora/corda.py lines ~102 and ~163, and src/peft/tuners/lora/loraga.py line ~101) call torch.load on config-specified cache/covariance files without weights_only=True, bypassing peft\u0027s own safe-loading wrapper used elsewhere in the codebase.\u003c/p\u003e"
                }
              ],
              "value": "Hugging Face peft\u0027s LoRA-GA and CorDA initialization modules (src/peft/tuners/lora/corda.py lines ~102 and ~163, and src/peft/tuners/lora/loraga.py line ~101) call torch.load on config-specified cache/covariance files without weights_only=True, bypassing peft\u0027s own safe-loading wrapper used elsewhere in the codebase."
            }
          ],
          "metrics": [
            {
              "cvssV3_1": {
                "attackComplexity": "LOW",
                "attackVector": "NETWORK",
                "availabilityImpact": "HIGH",
                "baseScore": 8.8,
                "baseSeverity": "HIGH",
                "confidentialityImpact": "HIGH",
                "integrityImpact": "HIGH",
                "privilegesRequired": "NONE",
                "scope": "UNCHANGED",
                "userInteraction": "REQUIRED",
                "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
                "version": "3.1"
              }
            }
          ],
          "problemTypes": [
            {
              "descriptions": [
                {
                  "cweId": "CWE-502",
                  "description": "CWE-502",
                  "lang": "en",
                  "type": "CWE"
                }
              ]
            }
          ],
          "providerMetadata": {
            "dateUpdated": "2026-08-10T11:47:04.582Z",
            "orgId": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
            "shortName": "TuranSec"
          },
          "references": [
            {
              "tags": [
                "third-party-advisory"
              ],
              "url": "https://github.com/huggingface/peft"
            },
            {
              "tags": [
                "third-party-advisory"
              ],
              "url": "https://github.com/huggingface/peft/blob/main/src/peft/tuners/lora/corda.py"
            }
          ],
          "source": {
            "discovery": "EXTERNAL"
          },
          "title": "peft Unsafe Deserialization via torch.load() Without weights_only in LoRA-GA and CorDA Modules"
        }
      },
      "cveMetadata": {
        "assignerOrgId": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
        "assignerShortName": "TuranSec",
        "cveId": "CVE-2026-71281",
        "datePublished": "2026-08-05T12:26:27.376Z",
        "dateReserved": "2026-08-05T12:23:34.968Z",
        "dateUpdated": "2026-08-10T11:47:04.582Z",
        "state": "PUBLISHED"
      },
      "dataType": "CVE_RECORD",
      "dataVersion": "5.2"
    }
  }
}

FKIE_CVE-2026-71281

Vulnerability from fkie_nvd - Published: 2026-08-05 13:24 - Updated: 2026-08-26 17:13

Severity

8.8 (High) - CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H

Summary

References

	URL	Tags
	309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c	https://github.com/huggingface/peft
	309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c	https://github.com/huggingface/peft/blob/main/src/peft/tuners/lora/corda.py

Impacted products

	Vendor	Product	Version

JSON

To clipboard

{
  "affected": [
    {
      "affectedData": [
        {
          "defaultStatus": "unaffected",
          "product": "peft",
          "programFiles": [
            "src/peft/tuners/lora/corda.py",
            "src/peft/tuners/lora/loraga.py"
          ],
          "vendor": "huggingface",
          "versions": [
            {
              "lessThanOrEqual": "0.19.1",
              "status": "affected",
              "version": "0",
              "versionType": "semver"
            }
          ]
        }
      ],
      "source": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c"
    }
  ],
  "cveTags": [],
  "descriptions": [
    {
      "lang": "en",
      "value": "Hugging Face peft\u0027s LoRA-GA and CorDA initialization modules (src/peft/tuners/lora/corda.py lines ~102 and ~163, and src/peft/tuners/lora/loraga.py line ~101) call torch.load on config-specified cache/covariance files without weights_only=True, bypassing peft\u0027s own safe-loading wrapper used elsewhere in the codebase."
    }
  ],
  "id": "CVE-2026-71281",
  "lastModified": "2026-08-26T17:13:24.800",
  "metrics": {
    "cvssMetricV31": [
      {
        "cvssData": {
          "attackComplexity": "LOW",
          "attackVector": "NETWORK",
          "availabilityImpact": "HIGH",
          "baseScore": 8.8,
          "baseSeverity": "HIGH",
          "confidentialityImpact": "HIGH",
          "integrityImpact": "HIGH",
          "privilegesRequired": "NONE",
          "scope": "UNCHANGED",
          "userInteraction": "REQUIRED",
          "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
          "version": "3.1"
        },
        "exploitabilityScore": 2.8,
        "impactScore": 5.9,
        "source": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
        "type": "Secondary"
      }
    ],
    "ssvcV203": [
      {
        "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
        "ssvcData": {
          "id": "CVE-2026-71281",
          "options": [
            {
              "exploitation": "none"
            },
            {
              "automatable": "no"
            },
            {
              "technicalImpact": "total"
            }
          ],
          "role": "CISA Coordinator",
          "timestamp": "2026-08-05T15:51:18.837851Z",
          "version": "2.0.3"
        }
      }
    ]
  },
  "published": "2026-08-05T13:24:52.723",
  "references": [
    {
      "source": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
      "url": "https://github.com/huggingface/peft"
    },
    {
      "source": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
      "url": "https://github.com/huggingface/peft/blob/main/src/peft/tuners/lora/corda.py"
    }
  ],
  "sourceIdentifier": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
  "vulnStatus": "Deferred",
  "weaknesses": [
    {
      "description": [
        {
          "lang": "en",
          "value": "CWE-502"
        }
      ],
      "source": "309f9ea4-e3e9-4c6c-b79d-e8eb01244f2c",
      "type": "Secondary"
    }
  ]
}

GHSA-G7PC-47RC-WVWF

Vulnerability from github – Published: 2026-08-05 15:32 – Updated: 2026-08-05 15:32

Details

Hugging Face peft's LoRA-GA and CorDA initialization modules (src/peft/tuners/lora/corda.py lines ~102 and ~163, and src/peft/tuners/lora/loraga.py line ~101) call torch.load() on config-specified cache/covariance files without weights_only=True, bypassing peft's own safe-loading wrapper used elsewhere in the codebase. Because torch.load() without weights_only=True performs full pickle deserialization, loading a malicious cache or covariance file (e.g. a shared/downloaded LoRA-GA or CorDA cache) results in arbitrary code execution.

Severity

8.8 (High)


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

Show details on source website

JSON

To clipboard

{
  "affected": [],
  "aliases": [
    "CVE-2026-71281"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-502"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2026-08-05T13:24:52Z",
    "severity": "HIGH"
  },
  "details": "Hugging Face peft\u0027s LoRA-GA and CorDA initialization modules (src/peft/tuners/lora/corda.py lines ~102 and ~163, and src/peft/tuners/lora/loraga.py line ~101) call torch.load() on config-specified cache/covariance files without weights_only=True, bypassing peft\u0027s own safe-loading wrapper used elsewhere in the codebase. Because torch.load() without weights_only=True performs full pickle deserialization, loading a malicious cache or covariance file (e.g. a shared/downloaded LoRA-GA or CorDA cache) results in arbitrary code execution.",
  "id": "GHSA-g7pc-47rc-wvwf",
  "modified": "2026-08-05T15:32:20Z",
  "published": "2026-08-05T15:32:20Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-71281"
    },
    {
      "type": "WEB",
      "url": "https://github.com/huggingface/peft"
    },
    {
      "type": "WEB",
      "url": "https://github.com/huggingface/peft/blob/main/src/peft/tuners/lora/corda.py"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ]
}

RHSA-2026:65126

Vulnerability from csaf_redhat - Published: 2026-09-08 07:57 - Updated: 2026-09-18 15:15

Summary

Red Hat Security Advisory: RHOAI 2.25.11 - Red Hat OpenShift AI

Severity

Important

Scope

117 vulnerabilities in this advisory, including CVE-2026-71281.

CVE-2026-71281

A flaw was found in peft. The LoRA-GA and CorDA initialization modules within Hugging Face peft improperly call `torch.load()` without the `weights_only=True` parameter. This oversight allows for full pickle deserialization when loading a malicious cache or covariance file. A remote attacker could exploit this by tricking a user into loading a specially crafted file, leading to arbitrary code execution on the affected system.

8.8 (High)


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

Fixed: 4 products Known not affected: 205 products

https://access.redhat.com/errata/RHSA-2026:65126

Open the full advisory

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.

The MITRE ATT&CK techniques below are AI-generated suggestions, inferred from the description of the vulnerability by the CIRCL/vulnerability-attack-technique-classification-roberta-base model, served locally by ML-Gateway. They have not been verified by an analyst and are provided for guidance only.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.

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.

Action not permitted

CVE-2026-71281 (GCVE-0-2026-71281)

FKIE_CVE-2026-71281

GHSA-G7PC-47RC-WVWF

RHSA-2026:65126

Tags

Sightings

Nomenclature

Related by attack behaviour