You’re both correct. The legal system has absolutely no idea how to handle the copyright issues around using content for AI training data. It’s a completely novel issue. At the same time, the tech companies have a lot more money to litigate favorable interpretations of the law than the content companies.