« Volver al listado

CVE-2025-53002

Estado: AnalizadaCrítica (9.8)—

LLaMA-Factory is a tuning library for large language models. A remote code execution vulnerability was discovered in LLaMA-Factory versions up to and including 0.9.3 during the LLaMA-Factory training process. This vulnerability arises because the `vhead_file` is loaded without proper safeguards, allowing malicious attackers to execute arbitrary malicious code on the host system simply by passing a malicious `Checkpoint path` parameter through the `WebUI` interface. The attack is stealthy, as the victim remains unaware of the exploitation. The root cause is that the `vhead_file` argument is loaded without the secure parameter `weights_only=True`. Version 0.9.4 contains a fix for the issue.

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.

RCE remota sin autenticación (AV:N, PR:N, UI:N) en API web de LLaMA-Factory al cargar vhead_file sin `weights_only=True`, permitiendo ejecución de código arbitrario.

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

Privileged Account ManagementVulnerability ScanningRestrict Web-Based ContentNetwork SegmentationLimit Software InstallationLimit Access to Resource Over Network

Tecnologías afectadas (1)

Hiyouga — Llama-factory

CWE

Referencias

JSON original (NVD)

Mostrar
{
  "id": "CVE-2025-53002",
  "cveTags": [],
  "metrics": {
    "ssvcV203": [
      {
        "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
        "ssvcData": {
          "id": "CVE-2025-53002",
          "role": "CISA Coordinator",
          "options": [
            {
              "exploitation": "poc"
            },
            {
              "automatable": "no"
            },
            {
              "technicalImpact": "partial"
            }
          ],
          "version": "2.0.3",
          "timestamp": "2025-06-26T15:05:57.527009Z"
        }
      }
    ],
    "cvssMetricV31": [
      {
        "type": "Secondary",
        "source": "security-advisories@github.com",
        "cvssData": {
          "scope": "UNCHANGED",
          "version": "3.1",
          "baseScore": 8.3,
          "attackVector": "NETWORK",
          "baseSeverity": "HIGH",
          "vectorString": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:H",
          "integrityImpact": "LOW",
          "userInteraction": "NONE",
          "attackComplexity": "LOW",
          "availabilityImpact": "HIGH",
          "privilegesRequired": "LOW",
          "confidentialityImpact": "HIGH"
        },
        "impactScore": 5.5,
        "exploitabilityScore": 2.8
      },
      {
        "type": "Primary",
        "source": "nvd@nist.gov",
        "cvssData": {
          "scope": "UNCHANGED",
          "version": "3.1",
          "baseScore": 9.8,
          "attackVector": "NETWORK",
          "baseSeverity": "CRITICAL",
          "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H",
          "integrityImpact": "HIGH",
          "userInteraction": "NONE",
          "attackComplexity": "LOW",
          "availabilityImpact": "HIGH",
          "privilegesRequired": "NONE",
          "confidentialityImpact": "HIGH"
        },
        "impactScore": 5.9,
        "exploitabilityScore": 3.9
      }
    ]
  },
  "affected": [
    {
      "source": "security-advisories@github.com",
      "affectedData": [
        {
          "vendor": "hiyouga",
          "product": "LLaMA-Factory",
          "versions": [
            {
              "status": "affected",
              "version": "< 0.9.4"
            }
          ]
        }
      ]
    }
  ],
  "published": "2025-06-26T15:15:23.873",
  "references": [
    {
      "url": "https://drive.google.com/file/d/1AddKm2mllsXfuvL4Tvbn_WJdjEOYXx4y/view?usp=sharing",
      "tags": [
        "Exploit"
      ],
      "source": "security-advisories@github.com"
    },
    {
      "url": "https://github.com/hiyouga/LLaMA-Factory/commit/bb7bf51554d4ba8432333c35a5e3b52705955ede",
      "tags": [
        "Patch"
      ],
      "source": "security-advisories@github.com"
    },
    {
      "url": "https://github.com/hiyouga/LLaMA-Factory/security/advisories/GHSA-xj56-p8mm-qmxj",
      "tags": [
        "Exploit",
        "Third Party Advisory"
      ],
      "source": "security-advisories@github.com"
    }
  ],
  "vulnStatus": "Analyzed",
  "weaknesses": [
    {
      "type": "Secondary",
      "source": "security-advisories@github.com",
      "description": [
        {
          "lang": "en",
          "value": "CWE-94"
        },
        {
          "lang": "en",
          "value": "CWE-502"
        }
      ]
    }
  ],
  "descriptions": [
    {
      "lang": "en",
      "value": "LLaMA-Factory is a tuning library for large language models. A remote code execution vulnerability was discovered in LLaMA-Factory versions up to and including 0.9.3 during the LLaMA-Factory training process. This vulnerability arises because the `vhead_file` is loaded without proper safeguards, allowing malicious attackers to execute arbitrary malicious code on the host system simply by passing a malicious `Checkpoint path` parameter through the `WebUI` interface. The attack is stealthy, as the victim remains unaware of the exploitation. The root cause is that the `vhead_file` argument is loaded without the secure parameter `weights_only=True`. Version 0.9.4 contains a fix for the issue."
    },
    {
      "lang": "es",
      "value": "LLaMA-Factory es una librería de optimización para modelos de lenguaje grandes. Se descubrió una vulnerabilidad de ejecución remota de código en versiones de LLaMA-Factory hasta la 0.9.3 (incluida) durante su entrenamiento. Esta vulnerabilidad surge porque el `vhead_file` se carga sin las protecciones adecuadas, lo que permite a atacantes maliciosos ejecutar código malicioso arbitrario en el sistema host simplemente pasando el parámetro `Checkpoint path` malicioso a través de la interfaz `WebUI`. El ataque es sigiloso, ya que la víctima desconoce la explotación. La causa principal es que el argumento `vhead_file` se carga sin el parámetro seguro `weights_only=True`. La versión 0.9.4 contiene una solución para este problema."
    }
  ],
  "lastModified": "2026-06-17T09:37:27.313",
  "configurations": [
    {
      "nodes": [
        {
          "negate": false,
          "cpeMatch": [
            {
              "criteria": "cpe:2.3:a:hiyouga:llama-factory:*:*:*:*:*:*:*:*",
              "vulnerable": true,
              "matchCriteriaId": "B9CFDBF9-20D2-4910-BD79-B1F8F7F60071",
              "versionEndExcluding": "0.9.4"
            }
          ],
          "operator": "OR"
        }
      ]
    }
  ],
  "sourceIdentifier": "security-advisories@github.com"
}