A new legal analysis argues that training AI models on copyrighted text—such as news articles—falls under the same “transformative use” exception that shields The New York Times’ reporting from copyright claims. The piece explains how courts have historically treated the creation of derivative works that add new meaning or purpose as fair use, and applies that logic to machine‑learning datasets. It also examines recent court decisions and industry responses, weighing the potential benefits for AI development against concerns about the erosion of authors’ control and revenue. The article invites readers to consider whether the same protections that protect journalism should extend to the rapidly expanding field of AI training.
Training a model on copyrighted work is the same transformative use exception that protects The New York Times' journalism.
