US DOJ backs OpenAI’s right to train AI on New York Times articles without paying copyright fees
Executive summary: On September 3 2026, the US Department of Justice filed a brief supporting OpenAI's position that training its AI models on New York Times articles does not violate copyright or cause commercial harm, in the ongoing lawsuit where the NYT alleges unauthorized use of its content. The DOJ's stance could shape the legal precedent for AI training on copyrighted material, affecting licensing costs for AI developers and revenue models for news publishers.
Who is involved: United States Department of Justice, OpenAI, New York Times.
Likely next: The court will consider the DOJ brief as it decides on the NYT's claims; OpenAI may seek to dismiss the case, while the NYT may pursue further legal arguments or seek a settlement.
On September 3 2026 the U.S. Department of Justice filed a brief in the lawsuit brought by the New York Times against OpenAI, arguing that the company’s use of the newspaper’s articles to train its AI models does not cause commercial harm and does not violate copyright law. The brief notes that while a general responsibility requirement remains, the core finding aligns with OpenAI’s position that the training falls under permissible use. This filing reflects the Trump administration’s broader stance that AI developers should be allowed to leverage large‑scale text collections without paying licensing fees for the purpose of model development. The DOJ’s intervention could shift the legal benchmark for how copyrighted material is treated in AI training, potentially reducing the pressure on firms to negotiate costly data‑licensing agreements with publishers. In the near term, other content owners may intensify their own legal challenges or seek legislative clarification, while AI companies might accelerate the ingestion of publicly available archives. How courts balance the responsibility clause mentioned in the brief with the permissive view of training will shape future business models for both AI developers and traditional media outlets.
Timeline
- — IA : l’administration Trump défend le droit d’OpenAI à utiliser les archives du « New York Times » sans verser de droits d’auteur (Le Monde — Économie)
- — Trump administration sides with OpenAI in lawsuit against New York Times (The Guardian — Technology)
Analysis — what this means
Sectors affected
- AI model training data market
- News content licensing sector
Regulatory implications
- DOJ brief argues that training LLMs on publicly available articles does not violate US copyright law, reinforcing a fair‑use stance.
- The case may influence how the EU AI Act’s text‑and‑data‑mining exception is interpreted for AI training.
Historical parallels
- September 2 2026: Trump administration filed a brief supporting OpenAI in the same New York Times copyright lawsuit (Guardian).
Key entities
Sources
Open the full interactive case file on Beyond →