OpenAI, an American company that developed the artificial intelligence system ChatGPT, has confirmed the presence of a vulnerability that allowed some users to view the addresses of other users’ conversations.
The system was introduced for free use last year and can conduct conversations, answer questions, write songs, and complete press topics accurately.
The issue was reported by some users of social networking sites Twitter and Reddit, who shared pictures of chat logs that they claimed did not belong to them.
OpenAI has fixed the issue and disabled the automated chat program for a short time.
The company confirmed that users were unable to access actual chats, only their addresses.
However, concerns remain about this technology invading users’ privacy.
Every conversation that ChatGPT makes is stored in the chat history of the user, which can be reviewed later.
Some users noticed chats in their log that they did not have with the chatbot. The CEO of OpenAI called the vulnerability “horrific” and apologized for the inconvenience caused.
GOOGL 
AMZN 
MET 
T 
WPM 
DM 
SVM 
CMC 
HKDUSD=X 
BABA 