{"data":{"id":"b1f49136-cab8-4739-bba9-4192232aaa1b","title":"CVE-2026-127222: SGLang through 0.5.23, fixed in commit 7b5fb39, contains a missing authorization vulnerability in the POST…","summary":"SGLang through 0.5.23, fixed in commit 7b5fb39, has a missing authorization flaw (CVE-2026-127222) in the POST /load_lora_adapter_from_tensors endpoint. The endpoint lacks the @auth_level(AuthLevel.ADMIN_OPTIONAL) decorator that its sibling admin endpoints use, so unauthenticated attackers on the network can call it without an API key. Attackers can load a LoRA adapter from attacker-supplied serialized tensors to manipulate the served model, and combined with a separately documented deserialization bypass, achieve unauthenticated remote code execution as root.","solution":"Fixed in commit 7b5fb39.","labels":["security"],"sourceUrl":"https://nvd.nist.gov/vuln/detail/CVE-2026-127222","publishedAt":"2026-11-18T21:17:48.157Z","cveId":"CVE-2026-127222","cweIds":["CWE-862"],"cvssScore":"7.5","cvssSeverity":"high","severity":"high","attackType":["other"],"issueType":"vulnerability","affectedPackages":null,"affectedPackageNames":null,"affectedPackageRefs":null,"affectedVendors":[],"affectedVendorsRaw":["SGLang"],"classifierModel":"claude-haiku-5-5","classifierPromptVersion":"v4","summaryPromptVersion":"v2","headline":"SGLang missing authorization on POST /load_lora_adapter_from_tensors","headlinePromptVersion":"h1","cvssVector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:H/A:N","attackVector":"network","attackComplexity":"low","privilegesRequired":"none","userInteraction":"none","exploitMaturity":"unknown","epssScore":0,"epssCheckedAt":"2026-11-19T00:12:31.782Z","kevDateAdded":null,"advisoryAliases":["GHSA-m386-99gj-pq7p"],"affectedPackagesSource":null,"affectedPackagesCheckedAt":"2026-11-19T00:12:36.307Z","patchAvailable":null,"disclosureDate":"2026-11-18T21:17:48.157Z","fixReleasedAt":null,"daysToFix":null,"fixCheckedAt":null,"capecIds":["CAPEC-122"],"crossRefCount":0,"attackSophistication":"moderate","impactType":["confidentiality","integrity"],"aiComponentTargeted":"inference","llmSpecific":false,"classifierConfidence":0.95,"researchCategory":null,"atlasIds":null,"fixes":[]}}