OpenAI recently introduced a new ChatGPT desktop app for macOS alongside its new AI model, GPT-4o. Days after its launch, a developer named Pedro José Pereira revealed a severe security threat in the app. This threat allowed anyone to read ChatGP conversations on a MacBook. Now, OpenAI has rolled out a new update for the ChatGPT macOS app to fix the critical security flaw. Here’s all you need to know.
ChatGPT for macOS Users Must Update Their App Right Away
The Verge reported that Pereira shared an app he made with the publication. This app could show a user’s ChatGPT conversations on any machine with a click of a button. A critical flaw in the app made this possible, as ChatGPT for macOS was saving user conversations in plain text in a non-secure environment.
Once the security flaw was pointed out, OpenAI released an update claiming it encrypts the chat. The Verge reports that after updating to the latest version of ChatGPT for macOS, Pereira’s app no longer works.
According to Pereira’s Thread post, OpenAI offered the ChatGPT macOS through their website. This means that the company had opted out of Apple’s sandbox protection.
For the uninitiated, Apple’s sandbox is a proprietary security system that runs an app and stores its data in an isolated environment. This means no app could access any other part of the system without necessary permissions granted.
Not adhering to Apple’s sandboxing requirements means that anyone who has access to your system can obtain user details stored in the chats. However, your conversations are still available to OpenAI to train the ChatGPT models and for safety reasons.
