Get A Quote
Written by Cyber Threat Intelligence Unit on 23 March 2023

Bug revealing titles of user conversations forces OpenAI to temporarily shut down ChatGPT

OpenAI, the renowned AI company, temporarily shut down its popular chatbot, ChatGPT, after a serious malfunction occurred on Monday. The bug exposed titles of user conversations, revealing that some users' conversation history was shared with other ChatGPT users.

While users could delete their conversation history until the incident, they couldn't see the questions asked by others on ChatGPT. Social media users quickly reacted to the extent of the bug, with screenshots showing only the titles of each exchange. However, user identity and interests were not revealed by the leak.

This incident raises concerns about the privacy of ChatGPT users, as OpenAI has access to all conversations with the AI. Users must not disclose sensitive information to ChatGPT as it could lead to other people knowing the identity of the users. OpenAI has repeatedly warned users of the dangers of using ChatGPT.

Sam Altman, CEO and co-founder of OpenAI, confirmed the existence of the problem caused by "a bug in an open source library" on March 22, 2023, but stated that only a "small percentage of users" were affected. Despite this, OpenAI took ChatGPT offline for several hours, including the paid ChatGPT Plus offering, as soon as the malfunction was identified.

While access to ChatGPT was quickly restored, the startup disabled the conversation history option and made it inaccessible for several days. Paid subscribers and free version users were treated equally. This limited bug serves as a reminder that users must be cautious when using AI and sharing personal information with it.

Related Posts