Microsoft says its AI chatbot rarely reproduces news article content

Microsoft has stated in new legal filings that its AI assistant, Copilot, rarely reproduces even full sentences from news articles or books, let alone substantial portions that could substitute for original works. The company is currently facing copyright claims from publishers, including The New York Times, and book authors.

As part of the discovery process in the lawsuit, Microsoft provided 8.2 million Copilot chat logs to an expert hired by news publishers. The company claims these logs were specifically selected "because they hit on keywords implicating use of News Plaintiffs' websites, and therefore the most likely to contain News Plaintiffs' works."

Microsoft's analysis of these selected logs indicates that a very small fraction of conversations contained content that could potentially infringe on copyright. This disclosure comes amidst ongoing litigation where publishers allege that AI services use copyrighted material for training without proper authorization.

The company's submission aims to demonstrate that its AI technology does not extensively copy or distribute published content, a key point in the legal disputes surrounding copyright and generative AI.