OpenAI’s AI Model to Use Copyrighted Materials and Pay Peanuts

OpenAI is offering news publishers a couple of million to use their copyrighted news articles to train their AI models.

• OpenAI is reportedly offering news publishers annual licensing deals ranging from $1 million to $5 million for the use of copyrighted news articles in training AI models.
• Apple is also engaging in talks with news companies, offering substantial multiyear agreements worth $50 million to use their data for AI training.

OpenAI is offering news publishers annual licensing deals from $1 million to $5 million for the use of copyrighted news articles in AI training.

Ever since generative AI became widely available, the source of its AI training dataset has been causing controversy. At the heart of the matter is the copyrighted content used in the training. Even The New York Times had had enough and sued both Microsoft and OpenAI for copyright infringement.

Right before that, OpenAI struck a multiyear licensing agreement with publisher Axel Springer, the parent company of Politico and Business Insider, to use its news content in ChatGPT. The writers themselves, you know the people doing the actual journalistic work, were not consulted. But I’ll get to that in a minute.

Numbers are starting to come out now. OpenAI is offering these deals for $1 million to $5 million. Apple is also talking with news companies, but they are offering $50 million multiyear agreements to use these companies’ data to train their AI.

These numbers seem to be on par with non-AI licensing deals like Meta’s alleged $3 million a year for news content when it launched the Facebook News tab. Others are pointing at Google’s $1 billion commitment to news organizations in 2020.

However, these are different instances. In the first situation, the AI is not only trained on them but can also produce news articles using the same style as the news outlet it was trained on. In the second situation, the copyrighted article shows up on the news tab under the news outlet’s name. In Google News’s case, for example, the link redirects you to the official website (You can now find us on Google News, by the way).

The first is borderline theft and forgery while the second is just a marketing and advertising method.

Back to the little guys. Let’s face reality for a quick second. Anything you produce while working for a company and on company time is not yours. It’s theirs. I might have written what you are currently writing and it’s under my name, but I don’t own the rights to it. So technically, these big media companies are in the clear. But ethically…. Something’s amiss.

As for these AI companies, they are willing to pay the big media corporations and go about things the legal way. But when it’s independent writers, all bets are off? Meta’s copyright infringement is okay as long as it’s one of the little guys.

Licensing deals are the right thing to do. All I’m suggesting is that what’s good for the goose is good for the gander.

Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Intelligent Tech sections to stay informed and up-to-date with our daily articles.