According to PANews, OpenAI has pledged to protect user privacy and data security, outlining its current data protection measures. Users can manage their data settings, and ChatGPT Free and Plus users can choose whether they want to allow their data to be used for future model improvements. OpenAI has explicitly stated that it will not use data from the API, ChatGPT Enterprise Edition, and Team Edition clients to train models, nor will ChatGPT “ephemeral conversations” be used for model training.
OpenAI emphasizes that the purpose of typical training is to learn language and how to understand and respond, not to obtain private information. They do not actively collect personal information, nor do they use public Internet information to create personal profiles, serve advertisements, or sell user data. Every time a question is answered, the model creates new content, rather than storing or copying and pasting previous training information.
Furthermore, OpenAI is committed to minimizing the personal information used in model training. The training model refuses to process requests for private or sensitive information and attempts to avoid generating responses containing private or sensitive information.