{
 "id": "AIR-2026-0232",
 "slug": "cve-2026-105753-vllm-multimodal-ipc-cache-desync-denial-of-s",
 "title": "CVE-2026-105753: vLLM Multimodal IPC Cache Desync Denial of Service",
 "summary": "A cache desync vulnerability in vLLM's multimodal inference engine can occur after a request is rejected, causing subsequent requests using the same media hash to trigger assertions and service availability failures. The issue has been patched in newer versions.",
 "type": "attack_on_ai",
 "categories": [
  "infra_vuln"
 ],
 "severity": 2,
 "targets": [
  "framework_or_tooling"
 ],
 "products": [
  "vLLM"
 ],
 "incident_date": "2026-10-05",
 "incident_date_precision": "day",
 "first_reported": "2026-10-05",
 "sources": [
  {
   "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-105753",
   "publisher": "NVD",
   "title": "CVE-2026-105753: vLLM is an inference and serving engine for large language models.",
   "published_at": "2026-10-05"
  },
  {
   "url": "https://github.com/advisories/GHSA-ph3r-5jfg-f84f",
   "publisher": "GitHub Advisory Database",
   "title": "CVE-2026-105753: vLLM: Mirrored multimodal IPC caches desync after a rejected request — a later request reusing the same media hash trips a receiver…",
   "published_at": "2026-10-06"
  }
 ],
 "source_count": 2,
 "confidence": 0.95,
 "status": "published",
 "origin": "seed_nvd",
 "tags": [
  "CVE-2026-105753",
  "vllm",
  "dos"
 ],
 "created_at": "2026-10-18T06:33:31.967Z",
 "updated_at": "2026-10-18T06:33:31.967Z"
}