CVE-2026-114049vLLM before 0.31.0 Stale allowed_token_ids Mask via InputBatch.condense

Beschreibung

vLLM before 0.31.0 contains a cross-request mask leak in InputBatch.condense() that allows remote attackers to constrain other clients' outputs by leaving stale allowed_token_ids rows. Attackers can submit concurrent constrained generate requests, including via unauthenticated /invocations, so recycled batch slots apply their token allowlist to victims' unrestricted requests.

Daten: CVE.org, NVD, CISA KEV, FIRST EPSS, Ubuntu. Wiedergegeben wie in der Quelle, auf Englisch.

Betroffene Produkte und Versionen

Referenzen und Patches

• GitHub Security Advisory (GHSA-6cxc-2vcg-w5qc) vendor-advisoryVendor Advisory
• VulnCheck Advisory: vLLM before 0.31.0 Stale allowed_token_ids Mask via InputBatch.condense third-party-advisoryThird Party Advisory