In the long negotiation between those who create knowledge and those who systematize it, a new chapter is being written in American courts. Microsoft has stepped forward to contest the scale of its alleged use of New York Times journalism in its Copilot chatbot, drawing a legal line between training an AI on published work and actively reproducing that work for users. The case, joined by multiple publishers and quietly endorsed by the Trump administration on the side of the technology companies, asks a question that will shape the information economy for generations: who owns the raw material