{
  "id": "chatgpt-lockdown-mode-bypass-checkpoint-2026-09",
  "name": "ChatGPT Lockdown Mode bypass - HTML preview sandbox escape (Check Point)",
  "date_reported": "2026-09-18",
  "researcher": "Alexey Bukhteyev",
  "org": "Check Point Research",
  "sources": [
    {
      "url": "https://research.checkpoint.com/2026/chatgpt-lockdown-mode-bypass-phishing-malware-delivery-and-silent-data-exfiltration/",
      "title": "Check Point: ChatGPT Lockdown Mode Bypass - Phishing, Malware Delivery and Silent Data Exfiltration",
      "publisher": "Check Point"
    }
  ],
  "magicsword_coverage": [],
  "borderline": true,
  "borderline_note": "Researcher disclosure; OpenAI fixed Sept 18, 2026; no evidence of in-the-wild exploitation",
  "ai_role": "ai_agent_or_mcp_as_target",
  "ai_role_secondary": null,
  "ai_role_detail": "HTML preview sandbox allowed popups; nested iframe contentWindow.open escaped sandbox; enabled chat history theft, fake-update malware, credential phishing via malicious prompts/GPTs/shared chats",
  "technique_summary": "Check Point researcher Alexey Bukhteyev disclosed ChatGPT Lockdown Mode bypass allowing HTML preview sandbox escape via contentWindow.open from nested iframe. Attack vectors: (1) zero-click desktop chat history theft, (2) fake-update malware delivery pages, (3) credential phishing via malicious prompts, custom GPTs, or shared chat links. OpenAI patched vulnerability Sept 18, 2026. No evidence of in-the-wild exploitation; marked borderline as researcher disclosure only.",
  "initial_access": "Malicious prompt, custom GPT, or shared chat link with HTML preview exploit",
  "ai_tooling_and_models": [
    "ChatGPT",
    "ChatGPT Desktop",
    "Custom GPTs"
  ],
  "other_tooling": [
    "HTML preview sandbox",
    "Nested iframe escape"
  ],
  "victims_and_sectors": "ChatGPT users; theoretical risk only (fixed before in-wild use)",
  "regions": [
    "unknown"
  ],
  "attribution_or_cluster": "Researcher disclosure (Alexey Bukhteyev, Check Point); no threat actor; OpenAI patched Sept 18",
  "scale_and_metrics": "PoC only; patched Sept 18, 2026; no in-wild exploitation documented",
  "prevention_angle": "OpenAI patched Sept 18, 2026; users should update ChatGPT Desktop; avoid untrusted shared chat links; review custom GPT permissions; implement sandbox escape monitoring",
  "related_magicsword_projects": [],
  "source_notes": "Check Point blog; researcher disclosure only; OpenAI fixed Sept 18, 2026; no threat actor or in-wild exploitation documented",
  "last_reviewed": "2026-11-10",
  "iocs": {
    "domains": [],
    "ips": [],
    "hashes": [],
    "urls": []
  },
  "atlas_techniques": [
    {
      "id": "AML.T0051",
      "rationale": "Malicious prompts/GPTs/shared chats could deliver HTML preview exploits",
      "source_urls": [
        "https://research.checkpoint.com/2026/chatgpt-lockdown-mode-bypass-phishing-malware-delivery-and-silent-data-exfiltration/"
      ]
    },
    {
      "id": "AML.T0077",
      "rationale": "Sandbox escape enabled chat history theft (zero-click desktop)",
      "source_urls": [
        "https://research.checkpoint.com/2026/chatgpt-lockdown-mode-bypass-phishing-malware-delivery-and-silent-data-exfiltration/"
      ]
    },
    {
      "id": "AML.T0057",
      "rationale": "Credential phishing via escaped sandbox popup windows",
      "source_urls": [
        "https://research.checkpoint.com/2026/chatgpt-lockdown-mode-bypass-phishing-malware-delivery-and-silent-data-exfiltration/"
      ]
    }
  ],
  "attack_flow": {
    "nodes": [
      {
        "id": "craft-exploit",
        "title": "Craft HTML preview exploit",
        "description": "Attacker creates malicious prompt/GPT/shared chat with nested iframe contentWindow.open escape",
        "kind": "execution",
        "source_urls": [
          "https://research.checkpoint.com/2026/chatgpt-lockdown-mode-bypass-phishing-malware-delivery-and-silent-data-exfiltration/"
        ]
      },
      {
        "id": "deliver-link",
        "title": "Deliver via prompt/GPT/shared chat",
        "description": "User interacts with malicious prompt, custom GPT, or shared chat link",
        "kind": "access",
        "source_urls": [
          "https://research.checkpoint.com/2026/chatgpt-lockdown-mode-bypass-phishing-malware-delivery-and-silent-data-exfiltration/"
        ]
      },
      {
        "id": "escape-sandbox",
        "title": "Escape HTML preview sandbox",
        "description": "Nested iframe contentWindow.open escapes Lockdown Mode sandbox",
        "kind": "execution",
        "source_urls": [
          "https://research.checkpoint.com/2026/chatgpt-lockdown-mode-bypass-phishing-malware-delivery-and-silent-data-exfiltration/"
        ]
      },
      {
        "id": "exfil-or-phish",
        "title": "Chat history theft or credential phishing",
        "description": "Zero-click chat history exfiltration (desktop) or credential phishing popup",
        "kind": "impact",
        "source_urls": [
          "https://research.checkpoint.com/2026/chatgpt-lockdown-mode-bypass-phishing-malware-delivery-and-silent-data-exfiltration/"
        ]
      }
    ],
    "edges": [
      {
        "source": "craft-exploit",
        "target": "deliver-link"
      },
      {
        "source": "deliver-link",
        "target": "escape-sandbox"
      },
      {
        "source": "escape-sandbox",
        "target": "exfil-or-phish"
      }
    ]
  },
  "category": "attacks-on-ai",
  "attribution": {
    "actor_name": "Researcher disclosure only",
    "country": "Unknown",
    "motivation": "unknown",
    "attributed_by": "Check Point",
    "confidence": "low"
  },
  "ai_involvement_confidence": "confirmed",
  "detections": [
    {
      "name": "ChatGPT security monitoring",
      "format": "guidance",
      "notes": "Update ChatGPT Desktop to post-Sept-18-2026 version; monitor for sandbox escape attempts; review shared chat links before opening; audit custom GPT permissions; implement popup blocker; detect contentWindow.open from nested iframes; educate users on malicious prompt/GPT risks"
    }
  ],
  "timeline": {
    "first_seen": "2026-09-18",
    "last_updated": "2026-09-18",
    "updates": [
      {
        "date": "2026-09-18",
        "source": "Check Point / OpenAI",
        "summary": "Alexey Bukhteyev discloses Lockdown Mode bypass; OpenAI patches same day; no in-wild exploitation documented"
      }
    ]
  }
}