{
 "public_id": "AIR-2026-0170",
 "incident_id": "2026-07-zenity-agentforger",
 "date_disclosed": "2026-07-23",
 "date_occurred": null,
 "title": "AgentForger: one phishing link forges an attacker-controlled ChatGPT workspace agent with a victim's identity",
 "class": "responsible_disclosure",
 "attack_surface": "enterprise_assistant",
 "initial_vector": "malicious_link_csrf_style",
 "agent_capability_abused": [
  "agent_builder",
  "oauth_delegation",
  "identity_inheritance",
  "autonomous_action"
 ],
 "target_product": "ChatGPT Workspace Agents / Agent Builder",
 "target_vendor": "OpenAI",
 "autonomy_level": "fully_autonomous",
 "impact_realized": false,
 "impact_type": [
  "persistent_unauthorized_access",
  "data_exfiltration",
  "credential_theft",
  "internal_phishing"
 ],
 "blast_radius": "ChatGPT enterprise workspaces using Agent Builder",
 "cve": [],
 "taxonomy": {
  "owasp_asi": [
   "ASI01",
   "ASI04",
   "ASI03"
  ],
  "mitre_atlas": [],
  "labels": [
   "agent_forgery",
   "cross_site_agent_forgery",
   "insider_simulation",
   "one_click"
  ]
 },
 "defense_present": null,
 "defense_bypassed": true,
 "fix_shipped": "OpenAI removed the over-permissive URL parameter within 4 days",
 "primary_sources": [
  "https://zenity.io/company-overview/newsroom/company-news/zenity-labs-uncovers-agentforger-a-chatgpt-vulnerability"
 ],
 "verified_quote": "A single link could hijack OpenAI's ChatGPT Agent Builder to stand up an attacker-controlled AI agent with a real employee's access.",
 "confidence": "high",
 "notes": "Disclosed to OpenAI via Bugcrowd 2026-06-04; acknowledged in 1 day, fixed in 4. Technical detail split across labs.zenity.io 'AgentForger, Part 1: ChatGPT Cross-Site Agent Forgery' and 'Part 2: The Autonomous Insider'. A dozen PoC scenarios; no known real-world abuse.",
 "risk_category": "security_fraud",
 "risk_secondary": [
  "privacy"
 ],
 "risk_subcategory": "Credential & Secret Theft",
 "risk_sub_secondary": [
  "Unauthorized Access & Account Takeover"
 ],
 "ai_entity": "generative",
 "agency": "agent_acted",
 "agency_debatable": false,
 "agency_reason": "the forged agent acted with the victim's identity and delegated OAuth",
 "vector_family": "malicious_component",
 "vector_detail": "malicious_link_csrf_style",
 "guardrail_kind": [
  "unknown"
 ],
 "guardrail_outcome": "silent",
 "xrefs": {
  "aiid": [],
  "atlas": [],
  "cve": [],
  "ghsa": []
 },
 "primary_asi": "ASI01",
 "agent_framework": "unknown",
 "model_provider": "openai",
 "model_name": "",
 "detection_mechanism": "external_researcher",
 "tool_access": "unknown",
 "financial_loss_band": "unknown",
 "eai_category": "credential_theft",
 "eai_domain": "access_control_and_permissions",
 "eai_secondary": [
  "privilege_escalation",
  "data_exfiltration_channels",
  "human_manipulation"
 ],
 "eai_reason": "subcategory: Credential & Secret Theft",
 "asi_curated": [
  "ASI01"
 ],
 "asi_scope": "in_scope",
 "asi_provenance": "curated_plus_rule",
 "asi_reason": "curated: ASI01 | added ASI04 (vector_family=malicious_component), ASI03 (eai_category=credential_theft)",
 "submitted_by": "EAI Internal"
}