Background
The New York Times Company sued OpenAI, Inc. and Microsoft Corporation in December 2023, alleging that large-language models including GPT-4 and Copilot were trained on Times-owned journalistic content without authorisation and that the models and associated tools reproduce protected works verbatim or in derivative form. The complaint raised claims for direct and contributory copyright infringement, violation of the Digital Millennium Copyright Act (§ 1202), unfair competition, and trademark dilution. In April 2025, Judge Sidney Stein issued a comprehensive order granting in part and denying in part the defendants’ motions to dismiss: the court allowed core copyright and DMCA claims to proceed while dismissing certain trademark and state-law counts. Discovery is ongoing.
AI interaction
The pleadings and April 2025 order recognise that OpenAI’s GPT-series and Microsoft’s Copilot systems were trained on "millions... of copyrighted news articles" from The New York Times, raising questions about whether large-scale dataset ingestion and output reproduction by generative models can constitute infringement or removal of copyright-management information. The court held that the Times had plausibly alleged copying of expressive text and that determining “substantial similarity” and “transformative” use requires factual evaluation at trial rather than dismissal at the pleading stage. The case thus frames the emerging U.S. approach to applying traditional copyright doctrines to AI-trained text generators. (See, ECF No. 1 and 514).
Note/Update:
The case remains in active discovery. Recent filings, including a Declaration in Support of Motion (ECF No. 982, 6 Nov 2025), show continuing disputes over the scope of document production and preservation of large-language-model training materials. The matter proceeds toward a potential evidentiary hearing or trial.