{
 "public_id": "AIR-2025-0005",
 "incident_id": "2025-02-openai-operator-unauthorized-purchase",
 "date_disclosed": "2025-02-07",
 "date_occurred": "2025-02-07",
 "title": "OpenAI Operator completed an unauthorized $31.43 grocery purchase despite the stated confirmation safeguard",
 "class": "safety_failure",
 "attack_surface": "browser_agent",
 "initial_vector": "none_adversarial",
 "agent_capability_abused": [
  "web_form_submission",
  "checkout",
  "stored_payment_method"
 ],
 "target_product": "OpenAI Operator",
 "target_vendor": "OpenAI",
 "autonomy_level": "semi_autonomous",
 "impact_realized": true,
 "impact_type": [
  "financial_loss",
  "unauthorized_transaction"
 ],
 "blast_radius": "one user; safeguard applied to all Operator users",
 "cve": [],
 "taxonomy": {
  "owasp_asi": [
   "ASI10",
   "ASI02"
  ],
  "mitre_atlas": [],
  "labels": [
   "confirmation_gate_failure",
   "excessive_agency",
   "consumer_browser_agent"
  ]
 },
 "defense_present": "OpenAI's stated requirement for user confirmation before purchases",
 "defense_bypassed": true,
 "fix_shipped": "OpenAI acknowledged the failure and committed to improving safeguards",
 "primary_sources": [
  "https://incidentdatabase.ai/cite/1028"
 ],
 "verified_quote": "It reportedly bypassed OpenAI's stated safeguard requiring user confirmation before purchases. OpenAI acknowledged the failure and committed to improving safeguards.",
 "confidence": "medium",
 "notes": "AI Incident Database incident 1028. Sole cited report is a Washington Post article behind a paywall; the quote is from the AI Incident Database's own incident page, cross-checked against the official AIID data snapshot (backup-20260824). Small dollar value but an early clean example of a human-in-the-loop purchase gate failing open.",
 "risk_category": "security_fraud",
 "risk_secondary": [],
 "risk_subcategory": "Unauthorized financial commitment",
 "risk_sub_secondary": [],
 "ai_entity": "generative",
 "agency": "agent_acted",
 "agency_debatable": false,
 "agency_reason": "Operator completed the checkout itself",
 "vector_family": "none_autonomous",
 "vector_detail": "none_adversarial",
 "guardrail_kind": [
  "human_approval"
 ],
 "guardrail_outcome": "silent",
 "xrefs": {
  "aiid": [
   1028
  ],
  "atlas": [],
  "cve": [],
  "ghsa": []
 },
 "primary_asi": "ASI10",
 "agent_framework": "openai_operator",
 "model_provider": "openai",
 "model_name": "",
 "detection_mechanism": "unknown",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "unsafe_actuation",
 "eai_domain": "agent_behaviour",
 "eai_secondary": [],
 "eai_reason": "subcategory: Unauthorized financial commitment",
 "asi_curated": [],
 "asi_scope": "in_scope",
 "asi_provenance": "rule",
 "asi_reason": "ASI10 (no adversary + unsafe actuation with no attacker-supplied trigger); ASI02 (impact=tool_misuse)",
 "submitted_by": "EAI Internal"
}