A Flaw in ChatGPT’s Mac App Could Have Let Hackers Grab Sensitive Data

The Mac ChatGPT app checks digital signatures three layers deep to stop rogue code from reaching its core, but researchers at Objective-See found a script interpreter that would accept an untrusted script if it was spawned three times to satisfy those checks. OpenAI has patched the flaw, which required malware already on the machine and could have exposed chat logs and browser sessions.

Summary written with AI by HowTechMoves, based on reporting by Wired.