US Backs OpenAI In Copyright Fight

The legal fight between OpenAI and The New York Times has just become even more important for the future of artificial intelligence. The Trump administration has backed OpenAI’s fair-use argument in the closely watched copyright case over whether AI companies can train models on large volumes of news articles without first paying publishers.

The case has always been bigger than one newspaper and one AI company. It sits at the centre of a question that almost every media house, author, musician, artist, software developer and AI lab is now trying to answer: when does training become fair use, and when does it become uncompensated extraction?

The government filing supports the idea that using publicly available material to train AI systems can be protected by fair use, especially when the output is presented as transformative technology. The New York Times, meanwhile, argues that its journalism was used without permission to create products that could compete with publishers and weaken the economics of original reporting.

This is why the case matters to everyone watching the AI economy. We have already seen music labels sue Anthropic over song lyrics, while the broader AI copyright fight has moved from books and images into journalism, music, video and software. Every ruling now becomes part of the map for what AI companies can build and what creators can protect.

OpenAI will welcome the government support, but it does not settle the case. Courts still have to weigh whether the training use is transformative, whether the copied work was lawfully obtained, whether outputs can reproduce protected material and whether the market for the original work is harmed.

The timing is also uncomfortable for publishers because AI companies are becoming more aggressive about content deals. Some media firms have signed licensing agreements, while others are choosing litigation. At the same time, Google is courting Hollywood for AI licensing , showing that even the biggest AI players know permissioned content may become a competitive advantage.

For readers, the simple version is this: if OpenAI wins broadly, AI companies will have stronger legal cover to train on much of the open web. If The New York Times wins broadly, AI labs may have to pay more, license more and build models with tighter content controls.

The outcome could decide whether AI training remains a mostly open-web activity or becomes a licensing market dominated by the companies that can afford the biggest deals. That is why this case is not just about copyright law. It is about who gets paid in the AI era and who gets turned into raw material.