{
 "public_id": "AIR-2026-0109",
 "incident_id": "2026-05-gtig-llm-account-pooling",
 "date_disclosed": "2026-05-12",
 "date_occurred": null,
 "title": "PRC-nexus clusters UNC6201/UNC5673 pooled stolen Gemini, Claude and OpenAI accounts via relay proxies",
 "class": "in_the_wild",
 "attack_surface": "agent_framework",
 "initial_vector": "account_pooling_via_open_source_relay_proxies",
 "agent_capability_abused": [
  "api_key_relaying",
  "account_pooling",
  "policy_evasion"
 ],
 "target_product": "Claude-Relay-Service, CLI-Proxy-API fronting Gemini/Claude/OpenAI",
 "target_vendor": "multiple",
 "autonomy_level": "assistant",
 "impact_realized": true,
 "impact_type": [
  "service_abuse",
  "policy_circumvention",
  "cost_shifting"
 ],
 "blast_radius": "Multiple pooled provider accounts across Google, Anthropic and OpenAI",
 "cve": [],
 "taxonomy": {
  "owasp_asi": [
   "ASI03"
  ],
  "mitre_atlas": [],
  "labels": [
   "llmjacking",
   "account_pooling",
   "attribution_evasion"
  ]
 },
 "defense_present": null,
 "defense_bypassed": null,
 "fix_shipped": null,
 "primary_sources": [
  "https://cloud.google.com/blog/topics/threat-intelligence/ai-vulnerability-exploitation-initial-access"
 ],
 "verified_quote": "they employ 'Claude-Relay-Service' to aggregate multiple Gemini, Claude, and OpenAI accounts, enabling account pooling and cost-sharing",
 "confidence": "high",
 "notes": "The operational-security layer beneath most of the other records in this corpus: it is how actors keep using frontier agents after bans. Directly relevant to agent credential/token theft as a market.",
 "risk_category": "security_fraud",
 "risk_secondary": [
  "product_misuse"
 ],
 "risk_subcategory": "Credential & Secret Theft",
 "risk_sub_secondary": [],
 "ai_entity": "generative",
 "agency": "human_acted_on_output",
 "agency_debatable": true,
 "agency_reason": "humans pooled stolen accounts through relay proxies",
 "vector_family": "conventional_vulnerability",
 "vector_detail": "account_pooling_via_open_source_relay_proxies",
 "guardrail_kind": [
  "unknown"
 ],
 "guardrail_outcome": "silent",
 "xrefs": {
  "aiid": [],
  "atlas": [],
  "cve": [],
  "ghsa": []
 },
 "primary_asi": "ASI03",
 "agent_framework": "unknown",
 "model_provider": "multiple",
 "model_name": "",
 "detection_mechanism": "vendor_disclosure",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "credential_theft",
 "eai_domain": "access_control_and_permissions",
 "eai_secondary": [
  "policy_drift"
 ],
 "eai_reason": "subcategory: Credential & Secret Theft",
 "asi_curated": [],
 "asi_scope": "in_scope",
 "asi_provenance": "rule",
 "asi_reason": "ASI03 (eai_category=credential_theft)",
 "submitted_by": "EAI Internal"
}