CVE-2026-107288
Pydantic AI es un framework de agentes en Python para crear aplicaciones y flujos de trabajo con IA generativa. Desde la 1.77.0 hasta la 1.107.6 y la 2.44.0, web_fetch_tool local y la alternativa local de WebFetch comparan las entradas de blocked_domains con el nombre de host de una URL antes de que ambos valores se normalicen a la forma utilizada por getaddrinfo.
Leer descripción completaMostrar menos
Un modelo influido por el atacante puede utilizar una grafía IDNA equivalente, un separador de etiquetas no ASCII, una variación de mayúsculas y minúsculas o una etiqueta raíz final que se resuelva en un host bloqueado pero no coincida con la cadena configurada, haciendo que la aplicación obtenga ese host con sus propios privilegios. allowed_domains falla en modo cerrado para las grafías no coincidentes, y las protecciones de IP privadas y de metadatos en la nube siguen siendo eficaces. Este problema está corregido en las versiones 1.107.6 y 2.44.0.
Traducción automática del texto original de NVD (en inglés).
CVSS
- Versión: 3.1
- Vector: CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N
- Puntuación base: 3.7
Probabilidad de explotación (EPSS)
- Probabilidad de explotación en los próximos 30 días: 0.38%
- Percentil entre todas las CVEs puntuadas: 30
- Fecha de la puntuación: 13/10/2026
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.
- Explotación
T1190Exploit Public-Facing Applicationinitial access72 % - Impacto principal
T1090.004Domain Frontingcommand and control78 % - Impacto secundario
T1552.007Container APIcredential access65 %
AV:N sin UI:N sugiere T1190. La vulnerabilidad permite SSRF mediante evasión de filtros de dominio (bypass de blocked_domains con variantes IDNA), accediendo a metadatos o hosts internos con privilegios de la aplicación.
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
Tecnologías afectadas (1)
CWE
- CWE-918, CWE-1289
Referencias
- https://github.com/pydantic/pydantic-ai/commit/490335f8e2322e143a79337ddca9410e0176c812
- https://github.com/pydantic/pydantic-ai/commit/a9dab92099d0ef9d5d4aa34ccac8a6f1b0e51284
- https://github.com/pydantic/pydantic-ai/commit/c1f212a084cbfa0012f2044cdb4731d214b3b983
- https://github.com/pydantic/pydantic-ai/pull/8407
- https://github.com/pydantic/pydantic-ai/pull/8409
- https://github.com/pydantic/pydantic-ai/pull/8421
- https://github.com/pydantic/pydantic-ai/releases/tag/v1.107.6
- https://github.com/pydantic/pydantic-ai/releases/tag/v2.44.0
- https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-22h6-qm39-v87j
JSON original (NVD)
Mostrar
{
"id": "CVE-2026-107288",
"cveTags": [],
"metrics": {
"ssvcV203": [
{
"source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"ssvcData": {
"id": "CVE-2026-107288",
"role": "CISA Coordinator",
"options": [
{
"exploitation": "none"
},
{
"automatable": "no"
},
{
"technicalImpact": "partial"
}
],
"version": "2.0.3",
"timestamp": "2026-10-08T17:28:31.014859Z"
}
}
],
"cvssMetricV31": [
{
"type": "Secondary",
"source": "security-advisories@github.com",
"cvssData": {
"scope": "UNCHANGED",
"version": "3.1",
"baseScore": 3.7,
"attackVector": "NETWORK",
"baseSeverity": "LOW",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N",
"integrityImpact": "NONE",
"userInteraction": "NONE",
"attackComplexity": "HIGH",
"availabilityImpact": "NONE",
"privilegesRequired": "NONE",
"confidentialityImpact": "LOW"
},
"impactScore": 1.4,
"exploitabilityScore": 2.2
}
]
},
"affected": [
{
"source": "security-advisories@github.com",
"affectedData": [
{
"vendor": "pydantic",
"product": "pydantic-ai",
"versions": [
{
"status": "affected",
"version": ">= 1.77.0, < 1.107.6"
},
{
"status": "affected",
"version": ">= 2.0.0b1, < 2.44.0"
}
]
},
{
"vendor": "pydantic",
"product": "pydantic-ai-slim",
"versions": [
{
"status": "affected",
"version": ">= 1.77.0, < 1.107.6"
},
{
"status": "affected",
"version": ">= 2.0.0b1, < 2.44.0"
}
]
}
]
}
],
"published": "2026-10-08T16:17:04.147",
"references": [
{
"url": "https://github.com/pydantic/pydantic-ai/commit/490335f8e2322e143a79337ddca9410e0176c812",
"tags": [
"Patch"
],
"source": "security-advisories@github.com"
},
{
"url": "https://github.com/pydantic/pydantic-ai/commit/a9dab92099d0ef9d5d4aa34ccac8a6f1b0e51284",
"tags": [
"Patch"
],
"source": "security-advisories@github.com"
},
{
"url": "https://github.com/pydantic/pydantic-ai/commit/c1f212a084cbfa0012f2044cdb4731d214b3b983",
"tags": [
"Patch"
],
"source": "security-advisories@github.com"
},
{
"url": "https://github.com/pydantic/pydantic-ai/pull/8407",
"tags": [
"Patch",
"Vendor Advisory"
],
"source": "security-advisories@github.com"
},
{
"url": "https://github.com/pydantic/pydantic-ai/pull/8409",
"tags": [
"Patch",
"Vendor Advisory"
],
"source": "security-advisories@github.com"
},
{
"url": "https://github.com/pydantic/pydantic-ai/pull/8421",
"tags": [
"Patch",
"Vendor Advisory"
],
"source": "security-advisories@github.com"
},
{
"url": "https://github.com/pydantic/pydantic-ai/releases/tag/v1.107.6",
"tags": [
"Patch",
"Release Notes",
"Vendor Advisory"
],
"source": "security-advisories@github.com"
},
{
"url": "https://github.com/pydantic/pydantic-ai/releases/tag/v2.44.0",
"tags": [
"Patch",
"Release Notes",
"Vendor Advisory"
],
"source": "security-advisories@github.com"
},
{
"url": "https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-22h6-qm39-v87j",
"tags": [
"Mitigation",
"Vendor Advisory"
],
"source": "security-advisories@github.com"
}
],
"vulnStatus": "Analyzed",
"weaknesses": [
{
"type": "Secondary",
"source": "security-advisories@github.com",
"description": [
{
"lang": "en",
"value": "CWE-918"
},
{
"lang": "en",
"value": "CWE-1289"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 1.77.0 until 1.107.6 and 2.44.0, the local web_fetch_tool and the WebFetch local fallback compare blocked_domains entries with a URL hostname before both values are normalized to the form used by getaddrinfo. An attacker-influenced model can use an equivalent IDNA spelling, non-ASCII label separator, case variation, or trailing root label that resolves to a blocked host but does not match the configured string, causing the application to fetch that host with its own privileges. allowed_domains fails closed for unmatched spellings, and private-IP and cloud-metadata protections remain effective. This issue is fixed in versions 1.107.6 and 2.44.0."
}
],
"lastModified": "2026-10-13T14:35:05.370",
"configurations": [
{
"nodes": [
{
"negate": false,
"cpeMatch": [
{
"criteria": "cpe:2.3:a:pydantic:pydantic_ai:*:*:*:*:*:python:*:*",
"vulnerable": true,
"matchCriteriaId": "B5355402-6E86-4212-A795-993C09433A3C",
"versionEndExcluding": "1.107.6",
"versionStartIncluding": "1.77.0"
},
{
"criteria": "cpe:2.3:a:pydantic:pydantic_ai:*:-:*:*:*:python:*:*",
"vulnerable": true,
"matchCriteriaId": "7FEC7EC8-5520-4930-8853-83108092F8BF",
"versionEndExcluding": "2.44.0",
"versionStartIncluding": "2.0.0"
}
],
"operator": "OR"
}
]
}
],
"sourceIdentifier": "security-advisories@github.com"
}