{
 "public_id": "AIR-2024-0025",
 "incident_id": "2024-06-paypal-ai-assistant-false-decline",
 "date_disclosed": "2024-07-13",
 "date_occurred": "2024-06-19",
 "title": "PayPal's generative AI assistant told a customer a transaction had been declined when no such transaction existed",
 "class": "safety_failure",
 "attack_surface": "consumer_chatbot",
 "initial_vector": "none_autonomous_action",
 "agent_capability_abused": [
  "proactive_notification",
  "account_state_reporting",
  "conversation_initiation"
 ],
 "target_product": "PayPal generative AI assistant (beta)",
 "target_vendor": "PayPal",
 "autonomy_level": "assistant",
 "impact_realized": true,
 "impact_type": [
  "integrity_violation",
  "service_disruption"
 ],
 "blast_radius": "one reported user, but the false alert was volunteered before any user input, so any beta user could receive one",
 "cve": [],
 "taxonomy": {
  "owasp_asi": [
   "ASI02",
   "ASI09"
  ],
  "mitre_atlas": [],
  "labels": [
   "no_adversary",
   "fabricated_account_state",
   "unprompted_output",
   "financial_context",
   "support_cost_shifted_to_user"
  ]
 },
 "defense_present": "none; the assistant was in beta and labelled as such",
 "defense_bypassed": false,
 "fix_shipped": "unknown",
 "primary_sources": [
  "https://www.wkiri.com/today/?p=1051",
  "https://incidentdatabase.ai/cite/748"
 ],
 "verified_quote": "Given that this happened to me before I typed anything to the chatbot, you can bet it's happening to others.",
 "confidence": "high",
 "notes": "Small in consequence, kept for the shape. The assistant did not answer a question badly — it volunteered a fabricated fact about the user's own account balance state, unprompted, in a financial product where an account state assertion is the thing users act on. Reported by researcher Kiri Wagstaff. The stated efficiency case for the agent inverted: it generated work for the customer and for the human support representative who had to unpick it. AI Incident Database incident 748.",
 "risk_category": "reliability",
 "risk_secondary": [],
 "risk_subcategory": "Hallucination & Factual Inaccuracy",
 "risk_sub_secondary": [
  "Load & Resilience Failures"
 ],
 "ai_entity": "generative",
 "agency": "agent_acted",
 "agency_debatable": true,
 "agency_reason": "the assistant volunteered a false account state unprompted; output-is-the-harm, no tool call",
 "vector_family": "none_autonomous",
 "vector_detail": "none_autonomous_action",
 "guardrail_kind": [
  "none"
 ],
 "guardrail_outcome": "not_applicable",
 "xrefs": {
  "aiid": [
   748
  ],
  "atlas": [],
  "cve": [],
  "ghsa": []
 },
 "primary_asi": "ASI02",
 "agent_framework": "unknown",
 "model_provider": "unknown",
 "model_name": "",
 "detection_mechanism": "unknown",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "hallucination",
 "eai_domain": "agent_output_quality",
 "eai_secondary": [
  "uncontrolled_resource_consumption",
  "confused_deputy"
 ],
 "eai_reason": "subcategory: Hallucination & Factual Inaccuracy",
 "asi_curated": [],
 "asi_scope": "in_scope",
 "asi_provenance": "rule",
 "asi_reason": "ASI02 (impact=tool_misuse); ASI09 (eai_category=hallucination + reliance signal)",
 "submitted_by": "EAI Internal"
}