OpenAI releases fix for ChatGPT Mac app exposing user conversations
What's the story
OpenAI has addressed a security flaw within its recently launched ChatGPT macOS app, which stored user conversations in plain text on a PC. This issue was discovered and demonstrated by a Threads user named Pedro Jose Pereira Vieito, who created an app that could access and display these conversations immediately after they occurred. The problem with storing chats in plain text is that they can be easily accessed by hackers or malicious apps, making all interactions with ChatGPT readable.
Swift action
An update to encrypt chats released
Upon being alerted by The Verge about the security issue, OpenAI promptly released an update to encrypt the chats. "We are aware of this issue and have shipped a new version of the application which encrypts these conversations," said OpenAI spokesperson Taya Christianson. She further emphasized that "We're committed to providing a helpful user experience while maintaining our high security standards as our technology evolves."
Enhanced security
Effectively securing user conversations
Following the update, Vieito's app was unable to access the conversations, and they were no longer visible in plain text. However, unless users opt out, OpenAI may still review ChatGPT conversations for safety and to train its models. This privilege does not extend to unknown third parties that gain access and know where to look.