Newly unsealed court documents from an ongoing lawsuit filed by the New York Times against OpenAI and Microsoft reveal that a Microsoft Director of Applied Science described AI model training as the largest theft of labor in human history, shedding harsh light on internal corporate anxieties over copyright and fair use.
Internal corporate communications unsealed from court records show that major technology executives used stark, unvarnished language to describe their data collection practices, privately clashing with the public legal defense of fair use. The disclosures emerge from unredacted legal files tied to the litigation initiated by the New York Times against OpenAI and Microsoft three years ago, offering a rare glimpse into boardroom anxiety over mass data scraping.
Inside Microsoft and OpenAI: Private Admissions Versus Public Defense
According to legal briefs unsealed, Microsoft’s Director of Applied Science, Brent Hecht, did not mince words when evaluating how companies ingest copyrighted material.
Satya Nadella’s Testimony and Paywall Evasion Tactics
Beyond executive commentary, the unredacted files outline the mechanical methods utilized to build massive training datasets. The documentation details techniques involving the circumvention of paywalls, large-scale web scraping, and the deliberate stripping of copyright management notices from source material.
The Legal Battlefield Over Fair Use and Market Substitution
Internal acknowledgments that AI products substitute for original journalism feed directly into plaintiffs’ arguments against fair use protection.
The revelations coincide with intense legal and political activity. The administration filed an amicus brief supporting OpenAI’s fair use stance. Simultaneously, a class-action lawsuit was launched by 400 local and regional American newspapers, accusing Microsoft and OpenAI of systematic and deliberate theft of hundreds of thousands of articles.
Urgent Licensing Efforts and What Lies Ahead
Faced with mounting judicial scrutiny and the risk of adverse rulings, OpenAI has pursued commercial settlements. The organization has actively sought to regularize its position through publisher licensing agreements that generally range between 1 and 10 million dollars annually.
The unsealed documents will feature prominently in upcoming summary judgment briefs, which will dictate precisely which claims survive to reach trial. The immediate legal horizon centers on judicial determinations regarding whether internal corporate admissions will decisively overturn the defendants’ fair use protections.
