Microsoft has signed an agreement with the popular publishing house HarperCollins Publishers. The deal will see the tech giant use the publication's non-fiction book titles to train artificial intelligence or AI models.
Microsoft has not yet revealed which models will use HarperCollins books for training. The company does not plan to use this content to create new books without human authors, but will instead use high-quality data to make the models more intelligent and accurate, said a person involved in the matter, who spoke on condition of anonymity.
Microsoft did not comment on the matter.
HarperCollins confirmed to Bloomberg News that the company has contracted with an undisclosed AI technology company. HarperCollins said AI models will allow certain non-fiction book titles to be used to improve quality and performance.
However, their authors will have the option to participate or not.
Author Daniel Kibblesmith posted a screenshot on social media. As seen in the screenshot, the license deal offered $2,500 per book. He said in another post on Tuesday, 'If I was offered a billion dollars, I would probably sign the contract. I will do it for the amount of money, which will not force me to work anymore. Because that is the end goal of this technology.'
Tech companies use a variety of data from social media sites to news reports to train AI models. Companies like Microsoft are looking to high-quality text sources so that programs can answer more accurate questions.
OpenAI signed an agreement with News Corporation last May. This will allow OpenAI to use content from various News Corporation publications, including The Wall Street Journal, Barron's and MarketWatch. In addition, Excel has signed license agreements with publishers such as Springer SE, The Atlantic, Vox Media, DotDash Meredith Inc., Hearst Communications Inc. and Time Magazine.
Microsoft is also working with Reuters, Hearst and Axel Springer to train AI models.
Some publishers have complained against AI companies for collecting content without permission. The New York Times sued OpenAI and Microsoft for copyright infringement. Artificial intelligence startup 'Perplexity AI' has also faced similar lawsuits.
