Can a Dutch case about RSS teach us anything about AI copyright?