Recently unsealed court documents in the New York Times' case against OpenAI and Microsoft paint a damning picture of internal awareness. The companies' own documentation warned that AI data scraping was starting a "doom loop" that would damage the web, characterized the practice as the "largest theft of labor in human history," and admitted it made a "complete mockery of the idea of fair use."
Many of the most striking quotes come from Microsoft's Director of Applied Science, Brent Hecht. Microsoft has attempted to distance itself from Hecht's assertions, with spokesperson Alex Haurek calling them "one employee's individual perspective" that does not represent the company's views. But the internal Microsoft document itself states: "Our AI content strategy has started a 'doom loop' that will hurt the performance of our models and the entire web at the same time."
Satya Nadella admitted in testimony that chatbots have basically replaced search, removing the need to visit source websites directly. OpenAI's own media experts attributed the drop in referral traffic for sites like the Times directly to AI summaries, estimating search referrals may be down as much as 60 percent.
Perhaps most damaging: OpenAI internally acknowledged that GPT-4 "memorized a ton of data and therefore will be insanely good at regurgitation," despite claiming that preventing memorization was important to minimize copyright violations. The filing cites multiple examples of ChatGPT outputting long strings copied verbatim from articles in the Times, Mercury News, The Denver Post, and others.
The documents reveal a fundamental tension: Microsoft admitted that "LLMs are a product that destroys its own supply chain" because AI output substitutes for the very training data it depends on. Both companies knew the consequences and proceeded anyway in pursuit of what OpenAI's Greg Brockman called "gazillions" of dollars.
