Introducing OpenAI Data Partnerships, where we'll work together with organizations to build public and private datasets for training AI models.

Modern AI technology learns skills and aspects of our world – people, our motivations, interactions and how we communicate – making sense of the data it's trained on. In order to ultimately make AGI that is safe and useful for all of humanity, we would like AI models to have a deep understanding of all topics, industries, cultures, and languages, which requires as broad a training data set as possible.

Including your content can make AI models more useful by increasing their understanding of your domain. We already work with many partners who want to present data from their country or industry. For example, we recently partnered with Icelandic government and Miðeind ehf improve GPT-4's ability to speak Icelandic by integrating their datasets. We have also partnered with a non-profit organization Free law project, which aims to democratize access to legal understanding by incorporating their large collection of legal documents into AI training. We know there could be many more who also want to contribute to the future of AI research as they unlock the potential of their unique data.

Data partnerships aim to enable more organizations to help drive the future of AI and take advantage of the models that work best for them, by including the content they care about.