Here's an article that seems relevant to the topic:

"District Court Finds That Using Copyrighted Works to Train Large Language Models Is Fair Use"

https://www.finnegan.com/en/insights/ip-updates/district-cou...