Dark Mode Light Mode

OpenAI signs license agreement with Axel Springer for model training.

Many, if not most, generative AI software suppliers believe that fair use allows them to train AI models using copyrighted information collected from the internet, even if the rightsholders do not grant permission. However, other suppliers, such as OpenAI, are hedging their bets, perhaps because they are concerned about the result of pending related cases.

OpenAI today announced a partnership with Axel Springer, the Berlin-based publisher of publications such as Business Insider and Politico, to train its generative AI models on the publisher’s content and to add recent Axel Springer-published articles to OpenAI’s viral AI-powered chatbot ChatGPT.

It’s OpenAI’s second agreement with a news company, after its announcement that it will license portions of The Associated Press’ archives for model training.

Advertisement

In the future, ChatGPT users will get summaries of “selected” items from Axel Springer’s publications, including news that is ordinarily behind a paywall. The excerpts will include credit as well as links to the full articles.

In exchange, Axel Springer will get payments from OpenAI of an undetermined magnitude and frequency. The agreement is valid for many years, and although neither party promises exclusivity, Axel Springer says it will support the outlet’s current AI-driven enterprises “that build upon OpenAI’s technology.”

“We’re excited to have shaped this global partnership between Axel Springer and OpenAI—the first of its kind,” said Axel Springer CEO Mathias Döpfner in a scripted statement. “We want to explore the opportunities of AI-empowered journalism—to bring quality, societal relevance, and the business model of journalism to the next level.”

Publishers and generative AI providers have a tense relationship, with the former charging copyright infringement and the latter increasingly worried about generative models cannibalizing traffic. For example, Google’s new generative AI-powered search experience, known as SGE, has moved conventional search results links lower down search results pages, possibly decreasing traffic to such links by up to 40%.

Publishers are also concerned about suppliers training their models on material without pay agreements in place, especially in light of rumors that digital behemoths like Google are experimenting with AI tools to summarize news. Hundreds of news organizations are now employing coding to block OpenAI, Google, and others from scanning their websites for training data, according to a recent study.

Several media groups, including Getty Images, The Associated Press, the National Press Photographers Association, and The Authors Guild, sent an open letter in August pushing for more transparency and copyright protection in AI. The signatories encouraged authorities, among other things, to explore legislation requiring openness in training data sets and allowing media businesses to engage with AI model operators.

“[Current] practices undermine the media industry’s core business models, which are predicated on readership and viewership (such as subscriptions), licensing, and advertising,” according to the letter. “In addition to violating copyright law, the resulting impact is to meaningfully reduce media diversity and undermine the financial viability of companies to invest in media coverage, further reducing the public’s access to high-quality and trustworthy information.”

Keep Up to Date with the Most Important News

By pressing the Subscribe button, you confirm that you have read and are agreeing to our Privacy Policy and Terms of Use
Add a comment Add a comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Previous Post

Meta begins experimenting with Threads integration with ActivityPub.

Next Post

Meta Quest 2, 3, and Pro now include Xbox Cloud Gaming.

Advertisement