Microsoft executives call AI news scraping a harmful ‘doom loop’
Microsoft executives said the company's AI systems rely on news content without compensating publishers, calling the practice a harmful “doom loop,” according to internal documents unsealed in the New York Times lawsuit. Earlier court filings had the executives describing the large-scale scraping of paywalled articles as theft, while internal memos from Microsoft and OpenAI acknowledged that such data collection could damage publishers and might not qualify as fair use. The documents show the companies debated the impact on the broader content ecosystem but maintained that their training methods fall under fair use.