Researchers bypassed ChatGPT Lockdown Mode to exfiltrate chat history, now fixed
What happened
Check Point Research showed that a ChatGPT HTML preview kept popup permissions that let a nested frame escape its sandbox, defeating Lockdown Mode. In the desktop app this could pull out chat history without user action. OpenAI fixed it on 18 September.
Check Point Research, as reported by The Hacker News, found that popup permissions in ChatGPT's HTML preview sandbox let a nested frame escape it, which bypassed Lockdown Mode. In the desktop app the researchers describe silent, zero-click theft of chat history. The issue is not limited to the Electron-based desktop environment. No exploitation is reported in the sources opened.
How it happened
Leaky AI product: App flaw. A security flaw in an AI application, such as broken login or access checks, let someone reach data they should not have seen.
The AI involved was from OpenAI. The product was ChatGPT, a type of chatbot.
Impact
Data was exposed. Data involved: chat logs. There is no evidence that anyone actually accessed the data.
Severity is rated Low, calculated from how far the incident reached and how sensitive the data was.
Was the AI at fault?
No. The failure was in the systems, settings or people around the AI, not in the model's behaviour.
Status and timeline
Confirmed by the affected organization, the AI vendor, a regulator or a named security research firm. This was a research finding: no real victims are known.
It became public on 5 Nov 2026.
OpenAI addressed the issue on 18 September after responsible disclosure.