Filtered by tag: check point research × Clear
Hackers Could Bypass ChatGPT Lockdown Mode to Steal Conversations Without Users Knowing
Matched: medical
Check Point Research found a flaw letting attackers bypass ChatGPT's Lockdown Mode via a hidden popup trick in HTML code previews. Exploiting nested iframes, attackers could steal full chat histories, display fake update windows to push malware, or capture login credentials. In the ChatGPT Classic desktop app, no user interaction was needed. OpenAI removed the popup permissions on September 18, 2026, fully closing the vulnerability.