Hardcoded trust_remote_code=True in vllm-project/vllm Bypasses User Security Control_CVE-2026-4944

{
    "lastseen": "",
    "description": "vllm-project/vllm version 0.14.1 contains a vulnerability where the `trust_remote_code=True` parameter is hardcoded in two model implementation files (`vllm/model_executor/models/nemotron_vl.py` and `vllm/model_executor/models/kimi_k25.py`). This bypasses the user's explicit `--trust-remote-code=False` setting, enabling remote code execution via malicious HuggingFace model repositories. This issue is an incomplete fix for CVE-2025-66448 and CVE-2026-22807, as it affects separate code paths in model implementation files. Deployments loading NemotronVL or KimiK25 models are particularly impacted.",
    "published": "2026-05-28T18:04:05.371Z",
    "modified": "2026-05-28T18:04:05.371Z",
    "type": "cve",
    "title": "Hardcoded trust_remote_code=True in vllm-project/vllm Bypasses User Security Control",
    "source": "@huntr_ai",
    "references": "https://huntr.com/bounties/97f706f7-a852-49b2-a4eb-76811e611daf",
    "id": "CVE-2026-4944",
    "bulletinFamily": "",
    "cwe": [
        "CWE-22"
    ],
    "cvelist": null,
    "sourceData": "vllm-project vllm-project/vllm unspecified",
    "sourceHref": "",
    "cvss": {
        "score": 8.8,
        "severity": "HIGH",
        "vector": "CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
        "version": "3.0"
    },
    "cvss2": [],
    "cvss3": {
        "version": "",
        "vectorString": "",
        "baseScore": 0,
        "baseSeverity": "",
        "attackVector": "",
        "attackComplexity": "",
        "privilegesRequired": "",
        "userInteraction": "",
        "scope": "",
        "confidentialityImpact": "",
        "integrityImpact": "",
        "availabilityImpact": "",
        "cvssV3": {
            "version": "",
            "vectorString": "",
            "baseScore": 0,
            "baseSeverity": "",
            "attackVector": "",
            "attackComplexity": "",
            "privilegesRequired": "",
            "userInteraction": "",
            "scope": "",
            "confidentialityImpact": "",
            "integrityImpact": "",
            "availabilityImpact": ""
        }
    },
    "href": "",
    "category_name": "CVE",
    "post_link": "",
    "product": "vllm-project/vllm",
    "version": "0.14.1",
    "vendor": "vllm-project",
    "ai_description": "Hardcoded trust_remote_code=True enables remote code execution via malicious HuggingFace model repositories",
    "ai_severity": "High",
    "ai_vendor": "vllm-project",
    "ai_product": "vllm",
    "ai_version": "0.14.1",
    "ai_score": 8.8
}