Microsoft has entered into an agreement with HarperCollins that will give the software giant access to content from the publisher to train its AI models. As per the terms of the agreement, the Redmond-based company will only have access to non-fiction titles.

The deal will reportedly allow Microsoft to improve an as-yet-unannounced AI model by training it on licensed nonfiction books.

Use precise geolocation data and actively scan device characteristics for identification. This is done to store and access information on a device and to provide personalised ads and content, ad and content measurement, audience insights and product development. List of Partners (vendors)

According to a statement HarperCollins gave to 404 Media, the agreement protects authors’ “underlying value of their works and our shared revenue and royalty streams.” Author Daniel Kibblesmith posted screenshots of an email showing that he would be paid $2,500 if he allowed one of his books to be licensed.

HarperCollins has not responded to requests for comment.

Microsoft’s AI training deal with HarperCollins’ nonfiction works meets author resistance over licensing terms.

The HarperCollins publishing company made a deal with an unnamed company, and now it wants authors to license their books to train AI.

Microsoft and HarperCollins have agreed to a licensing deal, according to Bloomberg, allowing the tech giant to train its models on books.

Authors are voicing concerns after a major book publisher offered payments in exchange for permission to use their books to train artificial intelligence.