ChatGPT has been reinstated in Italy said OpenAI and Italy’s data protection authority, after OpenAI resolved the concerns raised by the authority.
OpenAI said in a post OpenAI’s language models, such as ChatGPT, are developed using three primary sources of information: publicly available information on the internet, licensed third-party information, and information provided by users or human trainers. This article focuses on the first source of information, publicly available information, and how it is collected and used in compliance with privacy laws. The models do not contain or store copies of the information they learn from, and they do not have access to training information after they have learned from it.
Although a large amount of data on the internet relates to people, OpenAI does not actively seek out personal information to train its models. Personal information may incidentally be included in training information, but OpenAI will not use it to build profiles about people, contact them, advertise to them, sell to them, or sell the information itself. The purpose of using training information is solely to help the models learn about language and how to understand and respond to it.
To understand how OpenAI collects and uses information from users of its services, including how to opt out of having ChatGPT conversations used to help teach the models, users should refer to OpenAI’s Privacy Policy and help center article. The models are developed using publicly available information, licensed third-party information, and information provided by users or human trainers, and the information is used solely to help the models learn about language.
