Microsoft has formed a partnership with HarperCollins, a division of News Corp., to utilize select nonfiction titles to enhance its artificial intelligence models. This collaboration aims to improve the quality and performance of AI models by providing them with access to high-quality text. Sources reveal that Microsoft is using these books for a yet-to-be-announced AI project. The company has clarified it will not generate books without human authors and has refrained from further comments.
HarperCollins confirmed the agreement, emphasizing the limited and controlled use of specific nonfiction titles. The arrangement includes strict safeguards to protect authors’ rights and maintain shared royalty streams. Authors can choose whether to participate, ensuring their rights and interests are prioritized. HarperCollins reiterated its role in offering opportunities to authors while safeguarding the value of their work.
Tech companies frequently leverage diverse data sources, including articles and books, to train AI systems. Licensing agreements like these enable AI tools to become more accurate and efficient in responding to questions and delivering expertise.
News Corp. has already signed a deal with OpenAI, providing access to publications like the Wall Street Journal and MarketWatch. Microsoft and OpenAI have also collaborated with major publishers such as Hearst Communications, Time magazine, and Axel Springer. However, some publishers have raised concerns over unauthorized content use, leading to lawsuits like that of the New York Times against OpenAI and Microsoft, alleging copyright infringement.
