Microsoft and OpenAI Executives Privately Labeled AI Scraping as Theft

A January 2023 memo from Microsoft’s Director of Applied Science, Brent Hecht, characterized the unauthorized ingestion of human labor as the largest theft in history. This sentiment was echoed by internal OpenAI communications, where leadership admitted their chatbots pose an existential threat to the news organizations whose work fuels their systems. The documents detail how the companies bypassed paywalls, stripped copyright notices from training sets, and built massive datasets containing hundreds of thousands of articles from publications including The Times and the Daily News.

Evidence presented in the filing suggests the technology functions as a direct market substitute rather than a transformative tool. Microsoft’s own data revealed that the Copilot engine caused click-through rates for The New York Times to plummet by 93% compared to traditional search. CEO Satya Nadella testified that paywalled content should be licensed, noting he would have forced a model retrain had he known the extent of the scraping. Despite these internal warnings, the firms continue to lean on a fair-use defense in court, even as their own archives detail a systematic, deliberate effort to extract and repurpose copyrighted information at an unprecedented scale.