OpenAI offering media outlets as little as $1 million to use news articles for AI models: report

Photo of author

By Dan Sears

OpenAI is offering paltry payouts of as little as $1 million to media outlets in exchange for using their news articles to train the models that power ChatGPT — and the negotiations are playing out even as the company faces a major copyright infringement lawsuit from the New York Times.

The proposed licensing fees of between $1 million and $5 million are considered a “tiny amount even for small publishers,” The Information reported, citing two executives involved in recent talks with OpenAI.

OpenAI is said to be in negotiations with up to a dozen media outlets on potential licensing deals and has focused on global news operations, according to the report.

The firm previously inked a deal in December with publishing giant Axel Springer, the parent company of Politico and Business Insider, as well as a deal with the Associated Press last summer.

The financial terms of those deals were not disclosed at the time of their announcements. But executives told The Information that Axel Springer “is receiving tens of millions of dollars over several years.”

See also  DOJ close to filing massive antitrust suit against Apple: report

Two of OpenAI’s tech rivals — Google and Apple — are also reportedly seeking content deals with news publishers to fuel their AI products.

NYTimes and ChatGPT
News outlets are increasingly concerned about AI firms using their content. REUTERS

Apple is reportedly offering media firms more money in exchange for broader usage rights as it builds its own AI tools.

Google has reportedly trailed behind the other two companies in negotiations around AI usage but has existing relationships with some media outlets through its licensing deals for the Google News service.

The Post has reached out to OpenAI, Google and Apple for comment.

The negotiations are unfolding amid signs of growing concern among news outlets and other creatives of their use of content to train the so-called “large language models.” ChatGPT and other AI chatbots are trained on endless reams of internet data.

The New York Times sued OpenAI last week after negotiations broke down on a potential deal with the firm.

The NYTimes sued OpenAI last week. SARAH YENESEL/EPA-EFE/Shutterstock

The Gray Lady said OpenAI and its chief backer Microsoft had used “millions” of copyrighted articles to build their AI tools “without permission or payment.”

See also  Apple Q4 sales fall as big drop in Mac, iPad revenue offsets rise in iPhone sales

The lawsuit included specific instances in which chatbots regurgitated the Times’ content – including articles about Guy Fieri and investigative stories about Apple and the New York City taxi industry.

The newspaper “seeks to hold them responsible for the billions of dollars in statutory and actual damages,” according to the complaint.

Last week, OpenAI said it was “surprised and disappointed” that the New York Times had filed a lawsuit and was still “hopeful that we will find a mutually beneficial way to work together, as we are doing with many other publishers.”

“We respect the rights of content creators and owners and are committed to working with them to ensure they benefit from AI technology and new revenue models,” the statement said.

Rate this post

Leave a Comment