Check Point researchers found that popup handling in ChatGPT could bypass restrictions on content previews, allowing chat data and messages to be extracted without user action.
The flaw could affect more than the desktop app and bypassed the optional Lockdown Mode, designed to reduce the risk of conversation data extraction.
OpenAI fixed the issue on September 18, 2026, after responsible disclosure.
The issue was related to the way ChatGPT displayed content previews. Check Point found that handling pop-up windows could bypass the limitations of these previews.
The issue could also be used to display deceptive messages, such as a fake ChatGPT update notification. To download malicious software through such a message, someone would have to click on a link. The hidden copying of chat data, on the other hand, could be done without any clicks.
According to Check Point, the issue was not limited to the ChatGPT desktop app. OpenAI patched it on September 18, 2026, after the researchers duly notified the company.
>Stay connected with Hacks.gr
Set us as a preferred source on Google and find us on social media.
Preferred source
↗
end of article
Top comments (0)