• by Laurie Sullivan  @lauriesullivan, December 27, 2023

Several months after Microsoft made a commitment to legally back users of its products infused by artificial intelligence (AI) training data, The New York Times on Wednesday announced it had filed a lawsuit against Microsoft and OpenAI, the creator of ChatGPT.

The NYT copyright infringement lawsuit, which the publication could use as a bargaining chip for a long-term content deal, claimed that OpenAI copied millions of its articles to train the large language models (LLMs) that power ChatGPT and Microsoft Copilot without permission.

The copyright fight is just getting started. It may have begun with books, authors, musicians, and artists. Now in the hands of publishers, the media industry will start to see lucrative deals inked. Companies borrowing the data to train models have largely thought of it as using the information based on the U.S. "fair use" doctrine. 

Training data has become increasingly valuable. After all, Microsoft and Google integrate the information from publishers into their respective products for a variety of reasons, including serving up snippets into Google Search and Microsoft Bing search results, which compete with the New York Times for a source of information.

OpenAI and Microsoft should be liable for billions of dollars in damages, according to the NYT, which is seeking compensation likely either through the lawsuit or a licensing deal. This is something that publishers this year added to a list of grievances when considering journalistic content, images and other content to which search engines have access. 

Microsoft in October announced a Copilot Copyright commitment for customers using its products infused with AI. The commitment extended its existing intellectual property indemnity support to commercial Copilot services and builds on its previous AI Customer Commitments.

The commitment from Microsoft states that if a “third party sues a commercial customer for copyright infringement for using Microsoft’s Copilots or the output they generate, we will defend the customer and pay the amount of any adverse judgments or settlements that result from the lawsuit, as long as the customer used the guardrails and content filters" Microsoft has built into its products.

“The New York Times's complaint demonstrates the value of quality journalism to AI developers,” News/Media Alliance President and CEO Danielle Coffey stated in a release. “These companies repurpose and monetize news content, competing with the very industry they are benefiting from.”

A white paper analysis commissioned by the Alliance and released in October acknowledges the heavy reliance on journalistic and creative content by AI training models.  The findings, according to the Alliance, support a legal claim found in The New York Times’ complaint.

AI requires a lot of data to train models. The consequences for Microsoft, OpenAI, and others such as Google will include financial compensation to publishers. 

OpenAI signed a multiyear deal with Axel Springer earlier this month to use news stories to help train its AI models. Under the agreement, OpenAI will pay to use content to populate its AI models from Axel Springer publications, which include Politico and Business Insider in the U.S. and European properties Bild and Welt.

Apple reportedly has begun negotiations with major news publishers over permission to use their content as the technology company develops its own generative artificial intelligence (GAI) platforms.

The iPhone maker wants to secure multiyear deals worth at least $50 million to license the archives of news articles, according to the New York Times report, which cited people familiar with the discussions.