A major copyright dispute involving The New York Times, OpenAI and Microsoft has gained fresh attention following the release of previously sealed court documents. The documents contain internal discussions that the Times says show concerns within the technology companies about using news articles to develop artificial intelligence systems.
The New York Times has accused OpenAI of using millions of news articles to train its AI models without authorization. In a court filing made public on September 17, the newspaper pointed to comments attributed to Microsoft executive Brent Hecht, who reportedly described AI data collection in extremely critical terms.
According to the newly unsealed material, Microsoft and OpenAI employees had discussed the possible impact of AI systems on publishers and journalists. The filings also contain allegations concerning the collection of material from news websites and the use of large datasets for AI training.
The Times argues that its copyrighted articles were copied on a large scale and used to develop commercial AI products. The newspaper has also argued that AI-generated answers can potentially substitute for visiting original news websites, affecting publishers’ traffic and business models.
OpenAI and Microsoft have disputed the Times’ legal position. The companies have argued that using publicly available material to train AI models can qualify as fair use under US copyright law. Microsoft has also presented evidence from Copilot usage to argue that substantial reproduction of copyrighted material is uncommon.
The case is particularly significant for the journalism industry because AI companies rely heavily on large quantities of online information for training and improving their models. Publishers, meanwhile, are increasingly seeking licensing agreements and legal protections concerning the use of their content.
The legal battle began in 2023, when The New York Times sued OpenAI and Microsoft in federal court in Manhattan. The case has since become one of the most closely watched copyright disputes involving generative AI and news publishing.
There is also an important distinction between an allegation in court filings and a final judicial finding. The newly released documents show what the Times alleges and what internal communications reportedly contained; they do not by themselves establish that OpenAI or Microsoft violated copyright law.
The case could ultimately influence how AI companies obtain training data and how publishers protect and monetize journalism in the age of generative AI.