The Office's forthcoming report tackles the highest-stakes question in AI copyright law: the legal implications of training models on copyrighted works.
Docket links out to the original reporting rather than reproducing it — read the full story at the source below.