Microsoft exec called AI scraping 'largest theft of labor in human history'

Unredacted filings in the NYT lawsuit against OpenAI and Microsoft reveal internal admissions: data scraping for AI training was called 'theft,' and models an 'existential threat' to publishers. Microsoft data shows referrals to NYT from Copilot dropped to 93% compared to Bing.

• Microsoft: NYT referrals from Copilot dropped to 93% vs Bing
• OpenAI dataset found to contain 91,692 copies of NYT and other publishers' content
• Common Crawl dataset included over 2 million documents from nytimes.com
• Nadella said paid content should be licensed for training