Microsoft says Copilot rarely reproduces publishers’ material
Microsoft has submitted an analysis of 8.2 million Copilot chat logs in a copyright lawsuit brought by news publishers and authors. The company said the logs were selected using keywords associated with the publishers’ websites, making them more likely than ordinary conversations to contain material from those sites.
According to Microsoft, 59,545 conversations included at least 16 words matching news content used to train the AI model. The company also said an expert working for the Center for Investigative Reporting identified 51 cases of “substantial overlap” with CIR material.
In a separate authors’ lawsuit, Microsoft said an expert found only 24 Copilot responses containing at least 30 matching words across the 8.2 million conversations. Of 212 books examined, just 10 showed any matches, the company claimed. The New York Times, CIR and the Authors Guild had not immediately responded to requests for comment.
Written by Sakhanda Wire AI from reporting by The Verge. Facts, quotes and full context are in the original article below.