A U.S. court rules that using copyrighted data to train AI without permission does not qualify as fair use, setting a significant precedent for AI copyright disputes.
