« Volver al listado

CVE-2026-80047

Estado: Pendiente de análisisAlta (7.8)—

A vulnerability in Hugging Face Transformers (versions 4.57.0 to 5.16.1) allows remote Python files to be written to local disk without user consent when using GenerativePreTrainedModel.load_custom_generate(). The function fetches and caches a remote module file before performing the required trust_remote_code consent check, inverting the security model enforced by other code-loading paths (such as AutoConfig, AutoModel, and AutoTokenizer).

Leer descripción completaMostrar menos

As a result, attacker‑controlled Python code from custom_generate/generate.py is copied into the user’s ~/.cache/huggingface/modules directory even if the user declines the trust prompt. Although execution is correctly gated, the file write is not reversible and can persist across sessions. This can lead to persistent, unauthorized files on disk and stale cache collisions where cached attacker code may later be executed during trusted model loads. The issue stems from an unconditional file write in dynamic_module_utils.py prior to any trust verification.

CVSS

Probabilidad de explotación (EPSS)

EPSS (Exploit Prediction Scoring System, de FIRST) estima la probabilidad de que una vulnerabilidad sea explotada en 30 días. Complementa a CVSS (impacto) y a CISA KEV (explotación confirmada).

🎯 Técnicas ATT&CK

Cómo se explota esta vulnerabilidad y qué consigue el atacante, en el lenguaje de MITRE ATT&CK.

AV:L + PR:L + escritura de archivos remotos sin consentimiento previo en caché local. Escalada mediante código malicioso persistente en ~/.cache que ejecuta en futuras cargas de modelos de confianza.

Inferido por nuestro agente de análisis a partir de la descripción oficial, el vector CVSS y la CWE, y comprobado por un supervisor. Puede contener errores.

🛡️ Mitigaciones ATT&CK que cubren estas técnicas

Update SoftwareVulnerability ScanningThreat Intelligence ProgramRestrict File and Directory PermissionsRemote Data StorageExecution Prevention

Tecnologías afectadas (1)

⚠ Inferidas por IA a partir de la descripción — NVD aún no ha analizado esta CVE; no son CPE verificados.

Huggingface — TransformersAI

CWE

Referencias

JSON original (NVD)

Mostrar
{
  "id": "CVE-2026-80047",
  "cveTags": [],
  "metrics": {
    "ssvcV203": [
      {
        "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
        "ssvcData": {
          "id": "CVE-2026-80047",
          "role": "CISA Coordinator",
          "options": [
            {
              "exploitation": "none"
            },
            {
              "automatable": "no"
            },
            {
              "technicalImpact": "total"
            }
          ],
          "version": "2.0.3",
          "timestamp": "2026-09-03T13:52:40.495489Z"
        }
      }
    ],
    "cvssMetricV31": [
      {
        "type": "Secondary",
        "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
        "cvssData": {
          "scope": "UNCHANGED",
          "version": "3.1",
          "baseScore": 7.8,
          "attackVector": "LOCAL",
          "baseSeverity": "HIGH",
          "vectorString": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H",
          "integrityImpact": "HIGH",
          "userInteraction": "NONE",
          "attackComplexity": "LOW",
          "availabilityImpact": "HIGH",
          "privilegesRequired": "LOW",
          "confidentialityImpact": "HIGH"
        },
        "impactScore": 5.9,
        "exploitabilityScore": 1.8
      }
    ]
  },
  "affected": [
    {
      "source": "cret@cert.org",
      "affectedData": [
        {
          "vendor": "Hugging Face",
          "product": "Transformers",
          "versions": [
            {
              "status": "affected",
              "version": "4.57.0",
              "versionType": "custom",
              "lessThanOrEqual": "5.16.1"
            }
          ]
        }
      ]
    }
  ],
  "published": "2026-09-01T14:17:41.943",
  "references": [
    {
      "url": "https://github.com/huggingface/transformers",
      "source": "cret@cert.org"
    },
    {
      "url": "https://kb.cert.org/vuls/id/456290",
      "source": "cret@cert.org"
    },
    {
      "url": "https://www.kb.cert.org/vuls/id/456290",
      "source": "af854a3a-2127-422b-91ae-364da2661108"
    }
  ],
  "vulnStatus": "Awaiting Analysis",
  "weaknesses": [
    {
      "type": "Secondary",
      "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
      "description": [
        {
          "lang": "en",
          "value": "CWE-273"
        },
        {
          "lang": "en",
          "value": "CWE-367"
        },
        {
          "lang": "en",
          "value": "CWE-494"
        }
      ]
    }
  ],
  "descriptions": [
    {
      "lang": "en",
      "value": "A vulnerability in Hugging Face Transformers (versions 4.57.0 to 5.16.1) allows remote Python files to be written to local disk without user consent when using GenerativePreTrainedModel.load_custom_generate(). The function fetches and caches a remote module file before performing the required trust_remote_code consent check, inverting the security model enforced by other code-loading paths (such as AutoConfig, AutoModel, and AutoTokenizer). As a result, attacker‑controlled Python code from custom_generate/generate.py is copied into the user’s ~/.cache/huggingface/modules directory even if the user declines the trust prompt. Although execution is correctly gated, the file write is not reversible and can persist across sessions. This can lead to persistent, unauthorized files on disk and stale cache collisions where cached attacker code may later be executed during trusted model loads. The issue stems from an unconditional file write in dynamic_module_utils.py prior to any trust verification."
    }
  ],
  "lastModified": "2026-09-23T18:17:08.843",
  "sourceIdentifier": "cret@cert.org"
}