OpenAI's ChatGPT Mac App: A Privacy Concern Addressed In a recent development, OpenAI's ChatGPT application for Mac was found to have a significant security vulnerability. The app was storing user conversations in plain text format, potentially exposing sensitive information to anyone with access to the device[1]. This flaw was discovered by Pedro José Pereira Vieito, who demonstrated how easily another application could access and reveal ChatGPT conversations immediately after they occurred. The lack of encryption raised concerns about user privacy and data protection[1]. Upon being notified of this issue, OpenAI acted swiftly. They promptly updated the ChatGPT app to encrypt the stored records, significantly enhancing the protection of users' conversations[1]. It's worth noting that while OpenAI retains the ability to review ChatGPT conversations for safety and model training purposes, the accessibility of these conversations in plain text posed a potential risk if exploited by unauthorized parties[1]. This incident serves as a reminder of the importance of robust security measures in AI applications. As AI technology continues to advance, maintaining user trust through stringent data protection practices remains paramount. Sources [1] OpenAI’s ChatGPT Mac app was storing conversations in plain text https://lnkd.in/ghcZzpMT
And Signal stores plain text encryption keys 😑
So the unencrypted storing of sensitive data was built by design and then only addressed by OpenAI when exposed? Shocking!
Funny ChatGPT didn’t find that vulnerability in itself…