When Giants Collide: The New York Times Takes OpenAI to Court

New York Times’ Groundbreaking Copyright Lawsuit Against OpenAI and Microsoft Advances

In December 2023, The New York Times launched a high-stakes lawsuit in the Southern District of New York, accusing OpenAI and its key investor, Microsoft, of large – scale copyright infringement. The suit alleges that the companies used “millions” of the Times’ articles – ranging from news reporting to opinion columns and how-to guides – to train their large language models without permission or compensation

The Times argues that OpenAI’s ChatGPT and Microsoft’s Bing Chat (also known as Copilot) “free-ride” on its journalism to build substitute products that harm the newspaper’s subscription, licensing, advertising, and affiliate revenues.he complaint presents multiple examples where AI models recite Times articles verbatim or in close paraphrase – behavior the lawsuit links to a phenomenon known as “memorization” or “regurgitation”.

Beyond simple copying, the lawsuit also identifies instances of “hallucinations,” where AI tools falsely attribute content to the Times – such as an imagined list of “15 most heart – healthy foods” falsely claimed to be sourced from a Times article. The suit seeks “billions of dollars in statutory and actual damages,” plus a permanent injunction to prevent further use of Times content and orders to destroy any infringing models or data

OpenAI’s response has cited fair use, arguing that training on publicly available web data is lawful. In January 2024, company representatives said they were “surprised and disappointed” by the lawsuit and emphasized their commitment to content creator rights and ongoing discussions with publishers

NYT Sues OpenAI and Microsoft

In April 2025, a pivotal ruling came when U.S. District Judge Sidney Stein denied most of OpenAI’s motion to dismiss the case. He allowed essential copyright infringement claims to stand, including both direct and indirect liability allegations – and rejected OpenAI’s attempt to exclude older works from the complaint. This decision cleared the path to discovery, where both parties will exchange documents, evidence, and expert testimony.

OpenAI Claims Fair Use Defense

A separate court order tied to the case has required OpenAI to indefinitely preserve even user – deleted ChatGPT conversations. Previously, deleted chats were purged after 30 days. OpenAI plans to appeal the order, expressing concern that it violates user privacy norms

Wider Context: AI, Journalism, and Regulation

This lawsuit is the most prominent challenge by a media organization against AI firms. It follows similar legal actions by authors like Sarah Silverman and the Authors Guild, and aligns with legislative moves in countries such as Canada and Australia to require AI firms to pay for news usage. The Times’ action highlights key questions about AI’s impact on journalism and the evolving concept of fair use.

What Comes Next

Discovery Phase: Evidence-sharing begins, including Times internal data and AI training records.
Summary Judgment: After discovery, both sides will file motions that could potentially end the case or move it to trial.
Potential Trial: If unresolved, the case could proceed to a jury—a landmark moment in copyright law and AI.

A victory for the Times could reshape AI data-use practices, pushing companies toward licensing agreements. A defeat might reinforce a broad conception of permissible AI training. Either way, the outcome will deeply influence the future of AI, journalism, and intellectual property law.