ChatGPT Lockdown Mode bypass enables conversation theft
Original title: Hackers Could Bypass ChatGPT Lockdown Mode to Steal Conversations Without Users Knowing
Nested iframes and popups in ChatGPT preview let attackers exfiltrate full chat histories silently in desktop app.
A technical flaw in ChatGPT's Lockdown Mode permits bypass through nested iframes and popup triggers. This enables silent theft of conversations and files in the desktop app via shared chats or malicious prompts. OpenAI fixed the issue by removing popup permissions in September 2026 after Check Point Research disclosed it.
Unconfirmed: No confirmed real-world attacks reported by Check Point or OpenAI
- Popup trick uses allow-popups and allow-popups-to-escape-sandbox settings.
- Desktop app leaks chats on normal interaction without clicks.
- Fake update windows and login pages could also be used.
- Shared chats trigger data exfiltration without account needed.
Impact: stealing full chat histories and potential credential theft
- 1.Do not click previews or update prompts in shared chats
- 2.Skip prompts from unknown sites and custom GPTs
- 3.Verify address bar before entering passwords after logout
From the original article
3 min read at cybersecuritynews.com
A flaw in ChatGPT let attackers slip past Lockdown Mode, the safety setting meant to stop data leaks. Check Point Research found that a hidden pop-up trick could copy a user’s full chat history, show fake windows, and push malware. In the ChatGPT Classic desktop app, simply chatting was enough to trigger the leak.
By Guru Baran · cybersecuritynews.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.