Microsoft says virtually nobody was grabbing NYT articles through its chatbot
Executive Take
This fight will decide how much legal risk companies take on when they deploy AI tools trained on copyrighted content. A ruling for Microsoft would weaken publishers' leverage over every AI vendor, not just Copilot.
Executive Summary
Microsoft filed new legal arguments against copyright claims from The New York Times and book authors over its Copilot chatbot. Microsoft reviewed 8.2 million Copilot chat logs, selected by publishers' own expert for keywords tied to NYT content, and says Copilot rarely reproduces even full sentences from articles or books.
Why It Matters
AI leaders should track this because the outcome shapes licensing costs and legal exposure for any enterprise using generative AI tools built on scraped content. Technology executives need to know whether "fair use" defenses like Microsoft's will hold up in court.
Bizquad Perspective
The real story isn't whether Copilot copies text, it's that publishers had to hire an expert to search 8.2 million logs and still found only a small fraction of hits, suggesting the copying claim may be weaker than the lawsuit implies.