Earlier this week, internet users were surprised by the outage of chatbot ChatGPTChatGPT. In reality, it was OpenAI that was forced to disable the service after a data leak shared on social networkssocial networks.
The chatbotchatbot saves the history of conversations presented in the form of tabs, each with a title that summarizes the exchanges. However, netizens found that the history shown was not their own, and some shared screenshots on Reddit or Twitter with titles like “emotional support therapy” or “development of Chinese socialism”. Fortunately, it was not possible to access the exchanges’ content, instead the site displayed an error. However, just being able to read the titles of other users’ tabs is a disturbing invasion of their privacy.
The case was confirmed on Twitter by Sam Altman, CEO of OpenAI. According to him, the leak was caused by a bug in an open-source function library that has since been fixed. However, this raises the question of the processing and security of the personal data collected by OpenAI, especially knowing that the chatbot can be used to write personal messages, create documents for work or simply discuss sensitive topics.. .