{
 "public_id": "AIR-2025-0008",
 "incident_id": "2025-02-openai-operator-github-issue-injection",
 "date_disclosed": "2025-02-17",
 "date_occurred": null,
 "title": "ChatGPT Operator hijacked by a GitHub issue to exfiltrate PII from authenticated Hacker News and Booking.com sessions",
 "class": "research_demo",
 "attack_surface": "browser_agent",
 "initial_vector": "indirect_prompt_injection_web_page",
 "agent_capability_abused": [
  "authenticated_session_reuse",
  "cross_site_navigation",
  "form_filling",
  "web_browsing",
  "authenticated_session_access"
 ],
 "target_product": "OpenAI Operator",
 "target_vendor": "OpenAI",
 "autonomy_level": "semi_autonomous",
 "impact_realized": false,
 "impact_type": [
  "data_exfiltration",
  "pii_disclosure"
 ],
 "blast_radius": "Operator research-preview users",
 "cve": [],
 "taxonomy": {
  "owasp_asi": [
   "ASI01",
   "ASI03"
  ],
  "mitre_atlas": [],
  "labels": [
   "indirect_prompt_injection",
   "pii_exfiltration",
   "confirmation_bypass",
   "authenticated_session_abuse",
   "browser_agent"
  ]
 },
 "defense_present": "user monitoring, in-context confirmations, out-of-band confirmation when crossing site boundaries",
 "defense_bypassed": true,
 "fix_shipped": "none specific; OpenAI relies on human-in-the-loop confirmations",
 "primary_sources": [
  "https://embracethered.com/blog/posts/2025/chatgpt-operator-prompt-injection-exploits/"
 ],
 "verified_quote": "Operator cannot be fully trusted and these mitigations (which are necessary) will prevent creation of autonomous agents.",
 "confidence": "high",
 "notes": "Johann Rehberger. Injection hosted in a GitHub issue; Operator navigated to authenticated profile pages on Hacker News and Booking.com, read the user's email/address/phone and typed them into an attacker-controlled site. Key finding: typing text into a page rarely triggers a confirmation dialog, so exfiltration-by-typing evades the approval gate. Also reported as: 'ChatGPT Operator: prompt injection exfiltrates PII from authenticated sites'.",
 "risk_category": "privacy",
 "risk_secondary": [],
 "risk_subcategory": "PII Leakage",
 "risk_sub_secondary": [
  "Data Exfiltration via Context"
 ],
 "ai_entity": "generative",
 "agency": "agent_acted",
 "agency_debatable": false,
 "agency_reason": "Operator navigated authenticated sessions and exfiltrated PII",
 "vector_family": "indirect_prompt_injection",
 "vector_detail": "indirect_prompt_injection_web_page",
 "guardrail_kind": [
  "human_approval",
  "human_supervision"
 ],
 "guardrail_outcome": "silent",
 "xrefs": {
  "aiid": [],
  "atlas": [],
  "cve": [],
  "ghsa": []
 },
 "primary_asi": "ASI01",
 "agent_framework": "openai_operator",
 "model_provider": "openai",
 "model_name": "",
 "detection_mechanism": "external_researcher",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "sensitive_data_exposure",
 "eai_domain": "privacy",
 "eai_secondary": [
  "data_exfiltration_channels",
  "confused_deputy"
 ],
 "eai_reason": "subcategory: PII Leakage",
 "asi_curated": [
  "ASI01"
 ],
 "asi_scope": "in_scope",
 "asi_provenance": "curated_plus_rule",
 "asi_reason": "curated: ASI01 | added ASI03 (signal=identity)",
 "submitted_by": "EAI Internal"
}