Unsealed Court Filings Show Microsoft Exec Called AI Data Scraping “The Largest Theft of Labor in Human History
Newly unsealed filings say Microsoft and OpenAI knew their training practices threatened publishers, allegedly bypassed paywalls and removed copyright notices. OpenAI datasets reportedly included 91,692+ copies of works from major outlets and 2 million nytimes.com documents.
Newly unredacted court filings in The New York Times’ copyright lawsuit against OpenAI and Microsoft reveal that a senior Microsoft executive privately described the companies’ AI training practices as “the largest theft of labor in human history,” while internal documents show both companies were aware their AI products posed serious economic threats to the publishers whose content trained them.
The filings, unsealed in September 2026 as part of a lawsuit The Times originally filed three years ago, quote Microsoft’s Director of Applied Science, Brent Hecht, calling the scraping of copyrighted content “an astonishing theft of unprecedented proportions” in a January 2023 internal memo. Microsoft CEO Satya Nadella testified in a deposition that “anything that is paywalled should be licensed by anyone who wants to use it,” and said he would have required OpenAI to retrain its models had he known paywalled content had been scraped.
The documents detail how OpenAI and Microsoft allegedly obtained publisher content by bypassing paywalls undetected, building mass training datasets, and deliberately stripping copyright notices from training data. OpenAI’s mid-training datasets alone reportedly contain more than 91,692 copies of works from The New York Times, Daily News, and Center for Investigative Reporting. A Common Crawl-derived dataset included more than 2 million documents from nytimes.com. A joint initiative called Project Mango allegedly assembled a dataset containing at least 160,903 unique works from news publishers.
Internal communications also show OpenAI’s Head of ChatGPT, Nick Turley, describing publishers as facing an “existential threat” from products like ChatGPT, which he called “largely substitutive.” Microsoft’s own data showed its Copilot tool caused click-through rates to The New York Times’ domain to drop as much as 93% compared to traditional Bing search results.
These admissions could complicate OpenAI and Microsoft’s legal defense. The companies have argued their use of copyrighted material constitutes “fair use,” a legal doctrine that requires, among other factors, that the use does not harm the market for the original work. The Trump administration filed a brief earlier in 2026 supporting OpenAI’s fair use position. Neither OpenAI nor Microsoft responded to requests for comment.