Days after its launch on macOS, the ChatGPT app sparked security concerns due to chats being stored in plain text on users' computers. The Verge reported that all conversations with the chatbot were saved in plain text, raising alarm bells among users.
How the Issue Unfolded
A developer named Pedro Vieto discovered this security flaw and shared a video on Threads, revealing that ChatGPT was storing chats in plain text. Vieto's investigation shed light on the lack of app sandbox protections by OpenAI, allowing easy access to chat data.
Significance of the Issue
Storing conversations in plain text poses a significant security risk as it makes the data vulnerable to hacking attempts or unauthorized access. This means that personal conversations and sensitive information could be easily exposed.
Response from OpenAI
Upon being notified about the issue, OpenAI responded promptly. The company released an update to encrypt the stored conversations, ensuring that they are now secure and inaccessible to unauthorized parties. OpenAI's spokesperson, Taya Christianson, emphasized their commitment to user security and experience.
Ensuring Chat Security
While OpenAI states that they may use conversations to improve their models, users have the option to prevent their chats from being used for training purposes. By disabling this feature, users can enhance the security of their interactions with ChatGPT.
Source: The Verge
