News of the Cyber World

[kyonides]

ChatGPT Users’ Private Data Exposed Due to Open-Source Bug

ChatGPT, the artificial intelligence (AI) chatbot developed by OpenAI, was taken offline for emergency maintenance on March 20 due to a bug in an open-source library that triggered a data breach.

At the time, many ChatGPT users reported that they could see titles from other active user’s chat history. Some also reported seeing in their own history other people’s first messages of new conservations with the chatbot.

The bug has since been patched. Co-founder and CEO Sam Altman said on Wednesday on Twitter that OpenAI “felt awful” about the data breach.

On Friday, OpenAI provided an update on the incident, saying that even more private data from some users were exposed, including users’ payment information.

OpenAI stated that, upon deeper investigation, besides users seeing titles and conversations of others, the company “also discovered that the same bug may have caused the unintentional visibility of payment-related information of 1.2% of the ChatGPT Plus subscribers who were active during a specific nine-hour window.”

OpenAI did not disclose the exact number of paying accounts that were exposed.

...since its release as a free prototype to the public on Nov. 30, 2022, the company reported 100 million users in February.