For years, artificial intelligence advanced in a regulatory vacuum, its risks and concentrations of power treated as problems for another day. That era may be ending. Lawmakers in Washington, across party lines, are now openly reckoning with the question of whether a technology this consequential can be left to govern itself — and the answer, increasingly, is no. Yet the distance between political will and enacted law is vast, and the road ahead is lined with competing visions, institutional uncertainties, and a technology that refuses to slow down for the deliberations of democracy.