{"dataType": "CVE_RECORD", "dataVersion": "5.2", "cveMetadata": {"cveId": "CVE-2026-5241", "assignerOrgId": "c09c270a-b464-47c1-9133-acb35b22c19a", "state": "PUBLISHED", "assignerShortName": "@huntr_ai", "dateReserved": "2026-03-31T14:26:14.353Z", "datePublished": "2026-06-03T12:33:10.227Z", "dateUpdated": "2026-06-03T15:45:36.684Z"}, "containers": {"cna": {"title": "Policy Bypass in LightGlue Nested Config Resolution in huggingface/transformers", "providerMetadata": {"orgId": "c09c270a-b464-47c1-9133-acb35b22c19a", "shortName": "@huntr_ai", "dateUpdated": "2026-06-03T12:33:10.227Z"}, "descriptions": [{"lang": "en", "value": "A vulnerability in the LightGlue model loading path of huggingface/transformers version 5.2.0 allows an attacker-controlled model repository to execute arbitrary code during model initialization. The issue arises because the `trust_remote_code` parameter, intended to prevent remote code execution, is overridden by untrusted serialized configuration data in a nested code path. Specifically, when loading a LightGlue model using `AutoModel.from_pretrained()` with `trust_remote_code=False`, the `LightGlueConfig` reads the `trust_remote_code` value from the untrusted `config.json` file and propagates it into nested `AutoConfig.from_pretrained()` calls. This results in the execution of attacker-provided Python modules, even when the victim explicitly disables remote code execution. The vulnerability poses a high risk for environments such as API inference servers, research notebooks, CI/CD pipelines, and model evaluation workers, potentially leading to credential theft, lateral movement, or persistence/backdoor deployment."}], "affected": [{"vendor": "huggingface", "product": "huggingface/transformers", "versions": [{"version": "unspecified", "lessThan": "5.5.0", "status": "affected", "versionType": "custom"}]}], "references": [{"url": "https://huntr.com/bounties/ceb3ce1a-4c45-497a-b25e-cb9a7685e619"}, {"url": "https://github.com/huggingface/transformers/commit/676559d5022b74aaa0cee1cee0842b7f27c5320e"}], "metrics": [{"cvssV3_0": {"version": "3.0", "attackComplexity": "HIGH", "attackVector": "NETWORK", "availabilityImpact": "NONE", "confidentialityImpact": "HIGH", "integrityImpact": "HIGH", "privilegesRequired": "NONE", "scope": "CHANGED", "userInteraction": "REQUIRED", "vectorString": "CVSS:3.0/AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:N", "baseScore": 8, "baseSeverity": "HIGH"}}], "problemTypes": [{"descriptions": [{"type": "CWE", "lang": "en", "description": "CWE-829 Inclusion of Functionality from Untrusted Control Sphere", "cweId": "CWE-829"}]}], "source": {"advisory": "ceb3ce1a-4c45-497a-b25e-cb9a7685e619", "discovery": "EXTERNAL"}}, "adp": [{"references": [{"url": "https://huntr.com/bounties/ceb3ce1a-4c45-497a-b25e-cb9a7685e619", "tags": ["exploit"]}], "metrics": [{"other": {"type": "ssvc", "content": {"timestamp": "2026-06-03T15:45:14.993177Z", "id": "CVE-2026-5241", "options": [{"Exploitation": "poc"}, {"Automatable": "no"}, {"Technical Impact": "total"}], "role": "CISA Coordinator", "version": "2.0.3"}}}], "title": "CISA ADP Vulnrichment", "providerMetadata": {"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0", "shortName": "CISA-ADP", "dateUpdated": "2026-06-03T15:45:36.684Z"}}]}}

{}

{"cveTags": [], "descriptions": [{"lang": "en", "value": "A vulnerability in the LightGlue model loading path of huggingface/transformers version 5.2.0 allows an attacker-controlled model repository to execute arbitrary code during model initialization. The issue arises because the `trust_remote_code` parameter, intended to prevent remote code execution, is overridden by untrusted serialized configuration data in a nested code path. Specifically, when loading a LightGlue model using `AutoModel.from_pretrained()` with `trust_remote_code=False`, the `LightGlueConfig` reads the `trust_remote_code` value from the untrusted `config.json` file and propagates it into nested `AutoConfig.from_pretrained()` calls. This results in the execution of attacker-provided Python modules, even when the victim explicitly disables remote code execution. The vulnerability poses a high risk for environments such as API inference servers, research notebooks, CI/CD pipelines, and model evaluation workers, potentially leading to credential theft, lateral movement, or persistence/backdoor deployment."}], "id": "CVE-2026-5241", "lastModified": "2026-06-03T14:16:46.337", "metrics": {"cvssMetricV30": [{"cvssData": {"attackComplexity": "HIGH", "attackVector": "NETWORK", "availabilityImpact": "NONE", "baseScore": 8.0, "baseSeverity": "HIGH", "confidentialityImpact": "HIGH", "integrityImpact": "HIGH", "privilegesRequired": "NONE", "scope": "CHANGED", "userInteraction": "REQUIRED", "vectorString": "CVSS:3.0/AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:N", "version": "3.0"}, "exploitabilityScore": 1.6, "impactScore": 5.8, "source": "security@huntr.dev", "type": "Secondary"}]}, "published": "2026-06-03T14:16:46.337", "references": [{"source": "security@huntr.dev", "url": "https://github.com/huggingface/transformers/commit/676559d5022b74aaa0cee1cee0842b7f27c5320e"}, {"source": "security@huntr.dev", "url": "https://huntr.com/bounties/ceb3ce1a-4c45-497a-b25e-cb9a7685e619"}], "sourceIdentifier": "security@huntr.dev", "vulnStatus": "Received", "weaknesses": [{"description": [{"lang": "en", "value": "CWE-829"}], "source": "security@huntr.dev", "type": "Primary"}]}

{}

{"created": "2026-06-03T15:30:26.540882+00:00", "updated": "2026-06-03T15:30:26.540893+00:00", "vendors": []}