New Filings Deepen OpenAI Copyright Fight
Newly unredacted court filings add detail to the copyright dispute involving OpenAI, Microsoft and publisher content.
The copyright fight around generative AI is moving from broad legal arguments into more detailed evidence about how training data was obtained and how AI products could affect publishers.
What happened
Newly unredacted material in The New York Times' case against OpenAI and Microsoft revealed internal communications cited by the plaintiffs concerning AI training, publisher economics and the effect of AI-generated answers on news traffic.
The filing also includes allegations about the use of publisher content and methods for obtaining paywalled material.
Much of the newly public information comes from the plaintiffs' brief while some underlying exhibits remain sealed, meaning parts of the evidence still lack full public context. The allegations have not been finally adjudicated.
Why it matters
Copyright cases are increasingly turning on factual questions rather than abstract debates about whether AI training should qualify as fair use.
How content was collected, what companies understood about its provenance and whether AI products substitute for original publishing could all influence future litigation and licensing negotiations.
The bigger picture
The economics of AI and digital media are becoming tightly linked. Model developers want access to high-quality data, while publishers increasingly want compensation and control. The outcome of these disputes could shape the commercial rules around training data for years, affecting both frontier AI companies and smaller startups that depend on licensed datasets.