Microsoft argued that Copilot rarely copied the news content for New York Times case

Microsoft presented new evidence to defend itself in the copyright case launched by The New York Times and author associations. According to the claim of the company, Copilot, news articles and books **full phrases are rarely producing, creating large text pieces that can be crossed into the original content are never created.

The basic points for the legal application of Microsoft are:

• The company offered 8.2 million Copilot chat record in the scope of the discovery stage of the case to be examined by the specialist assigned by the publisher representatives
• Selected records were included in cases with keywords that are particularly high in the possibility of violation of broadcast copyright
• Microsoft argues that even results in this comprehensive sample do not have significant text replica
• Millions of records included in the study show that the company pays attention to copyrights

The case brings important questions about how to organize legal relationship between educational data sources with artificial intelligence models. Microsoft’s defense represents a new approach to prove that the output of AI systems does not contain copyright-right material. However, the case continues to be in a long process that will determine how the use of data on an unprecedented scale may affect copyright laws.