Internal documents from Microsoft and OpenAI, unsealed as part of a lawsuit led by The New York Times, reveal that the companies were aware their AI training practices would harm news organisations. Microsoft Director of Applied Science Brent Hecht repeatedly warned in internal communications that scraping news for AI model training constituted "an astonishing theft of unprecedented proportions" and possibly the "largest theft of labor in human history," directly contradicting the companies' public claims about fair use. OpenAI executives similarly acknowledged that publishers would face an "existential threat" from AI products designed to substitute for news platforms.

Data from Microsoft demonstrates the impact, with click-through rates for some news organisations dropping by 83–93 per cent, whilst others experienced 51–94 per cent declines. Internal documents describe a self-reinforcing "doom loop" in which declining news revenue reduces the quality and quantity of content available to train AI models, ultimately undermining the very tools these companies have built. Employees at both firms acknowledged in internal messages that users have no incentive to click news links when chatbots provide information directly, and OpenAI apparently developed a method to bypass The New York Times' paywall. News organisations assert they have compelling evidence of direct substitution and plan to proceed to trial, arguing that preserving human-created journalism is essential for both responsible AI development and society.