{
 "id": "AIR-2025-0146",
 "slug": "openai-model-bypassed-to-generate-fraudulent-receipts-and-pr",
 "title": "OpenAI Model Bypassed to Generate Fraudulent Receipts and Prescriptions",
 "summary": "Users bypassed safety restrictions in OpenAI's image generation model to produce fraudulent documents such as fake restaurant receipts and medical prescriptions. The generated material was used to demonstrate expense and medical forgery capabilities.",
 "type": "ai_enabled_attack",
 "categories": [
  "jailbreak",
  "other"
 ],
 "severity": 1,
 "targets": [
  "OpenAI"
 ],
 "products": [
  "ChatGPT"
 ],
 "incident_date": "2025-03-31",
 "incident_date_precision": "day",
 "first_reported": "2025-03-31",
 "sources": [
  {
   "url": "https://incidentdatabase.ai/cite/1018",
   "publisher": "AI Incident Database",
   "title": "OpenAI's 4o Model Allegedly Used to Generate Fake Receipts and Prescriptions",
   "published_at": "2025-03-31"
  }
 ],
 "source_count": 1,
 "confidence": 0.9,
 "status": "published",
 "origin": "seed_aiid",
 "tags": [
  "jailbreak",
  "fraud",
  "forgery"
 ],
 "created_at": "2026-10-18T06:33:31.967Z",
 "updated_at": "2026-10-18T06:59:35.797Z"
}