Dispatch
Microsoft's summary-judgment filing puts a number on Copilot regurgitation: 0.7% of 8.2M chats
Microsoft filed its motion for summary judgment on September 4 in the consolidated NYT and book-author copyright case, arguing Copilot rarely reproduces substantive passages. Of 8.2 million chat records analyzed, 59,545 (about 0.7%) contained at least 16 words in common with the plaintiffs' news content. The catch Microsoft states openly: the 8.2M logs were not a random sample, they were selected because they hit keywords implicating the plaintiffs' sites, so this is the worst-case slice, not the average conversation.
↳ Follow the thread