{
 "public_id": "AIR-2026-0187",
 "incident_id": "2026-08-wiz-litellm-ai-gateway-exploitation",
 "date_disclosed": "2026-08-06",
 "date_occurred": null,
 "title": "LiteLLM AI gateway: four security events in six months, including in-the-wild exploitation and MCP auth bypass",
 "class": "in_the_wild",
 "attack_surface": "agent_framework",
 "initial_vector": "unauthenticated_network_request",
 "agent_capability_abused": [
  "mcp_gateway",
  "model_proxying",
  "backend_credential_holding"
 ],
 "target_product": "LiteLLM AI gateway (incl. its MCP gateway)",
 "target_vendor": "BerriAI / LiteLLM",
 "autonomy_level": "semi_autonomous",
 "impact_realized": true,
 "impact_type": [
  "remote_code_execution",
  "unauthorized_access",
  "credential_theft",
  "cryptomining"
 ],
 "blast_radius": "Wiz observes LiteLLM in over a third of the cloud environments it monitors",
 "cve": [
  "CVE-2026-59822",
  "CVE-2026-42271",
  "CVE-2026-48710"
 ],
 "taxonomy": {
  "owasp_asi": [
   "ASI04",
   "ASI05",
   "ASI03"
  ],
  "mitre_atlas": [],
  "labels": [
   "ai_infrastructure",
   "auth_bypass",
   "mcp_gateway",
   "honeypot_telemetry",
   "in_the_wild"
  ]
 },
 "defense_present": null,
 "defense_bypassed": true,
 "fix_shipped": null,
 "primary_sources": [
  "https://www.wiz.io/blog/cloud-threat-highlights-h1-2026",
  "https://www.cisa.gov/news-events/alerts/2026/09/02/cisa-adds-seven-known-exploited-vulnerabilities-catalog"
 ],
 "verified_quote": "LiteLLM, an AI gateway present in over a third of the cloud environments we monitor, had four separate security events in six months: a supply-chain compromise, an SQL injection vulnerability exploited in the wild, a privilege escalation chain and an authentication bypass.",
 "confidence": "high",
 "notes": "Wiz Threat Research H1 2026 report. CVE numbers (CVE-2026-59822 MCP gateway auth bypass; CVE-2026-42271 MCP test-endpoint command injection; CVE-2026-48710 Starlette host-header bypass) come from Wiz search-result summaries and were not each individually re-quoted from the cited page - treat the CVE list as lower confidence than the quoted claim. Wiz also reports unauthenticated MCP endpoints across hundreds of environments. Update 2026-09-21: CISA added CVE-2026-59822 to the Known Exploited Vulnerabilities catalog on 2026-09-02 (alert 'CISA Adds Seven Known Exploited Vulnerabilities to Catalog', which also lists CVE-2026-48710 Starlette and CVE-2026-49869 Kestra). Related later records: 2026-08-exposed-ai-infrastructure-exploitation-campaign (the in-the-wild exploitation campaign, Microsoft and Wiz) and 2026-09-wiz-litellm-offguard-exposure (default-key exposure measurement plus CVE-2026-59821).",
 "is_umbrella": true,
 "umbrella_constituents": [
  "2026-08-berriai-litellm-llm-heist"
 ],
 "umbrella_reason": "Wiz summary spans four LiteLLM events; the LLM Heist event is separately cataloged.",
 "risk_category": "security_fraud",
 "risk_secondary": [],
 "risk_subcategory": "Insecure Code Generation",
 "risk_sub_secondary": [
  "Supply Chain Compromise",
  "Credential & Secret Theft",
  "Remote Code Execution & Sandbox Escape",
  "Unauthorized Access & Account Takeover"
 ],
 "ai_entity": "generative",
 "agency": "ai_as_target",
 "agency_debatable": false,
 "agency_reason": "unauthenticated exploitation of the LiteLLM gateway in the wild",
 "vector_family": "conventional_vulnerability",
 "vector_detail": "unauthenticated_network_request",
 "guardrail_kind": [
  "unknown"
 ],
 "guardrail_outcome": "silent",
 "xrefs": {
  "aiid": [],
  "atlas": [],
  "cve": [
   "CVE-2026-59822",
   "CVE-2026-42271",
   "CVE-2026-48710"
  ],
  "ghsa": []
 },
 "primary_asi": "ASI04",
 "agent_framework": "mcp_server",
 "model_provider": "unknown",
 "model_name": "",
 "detection_mechanism": "external_researcher",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "hallucination",
 "eai_domain": "agent_output_quality",
 "eai_secondary": [
  "credential_theft",
  "privilege_escalation"
 ],
 "eai_reason": "subcategory: Insecure Code Generation",
 "asi_curated": [
  "ASI04"
 ],
 "asi_scope": "in_scope",
 "asi_provenance": "curated_plus_rule",
 "asi_reason": "curated: ASI04 | added ASI05 (impact=code_exec), ASI03 (signal=identity)",
 "submitted_by": "EAI Internal"
}