According to PANews, OpenAI's official announcement promised to protect user privacy and data security, and introduced its current data protection measures. Users can manage data through settings, and ChatGPT free and Plus users can choose whether to allow their data to be used for future model improvements. OpenAI clearly stated that it will not use the data of API, ChatGPT Enterprise Edition and Team Edition customers to train models, and ChatGPT's "temporary conversations" will not be used for model training.
OpenAI emphasizes that the purpose of model training is to learn language and how to understand and respond, rather than to obtain private information. They do not actively collect personal information, nor do they use public Internet information to create personal profiles, place advertisements, or sell user data. Each time a question is answered, the model generates new content instead of storing or copying and pasting previous training information.
OpenAI is committed to reducing the personal information used to train models, training models to refuse to process requests for private or sensitive information, and trying to avoid generating responses that contain private or sensitive information.