China’s Open-Model Case Faces a New Global AI Alignment Test
Chinese AI developers’ use of downloadable model weights is moving a familiar U.S.-China technology contest into a harder governance question: who can set safeguards after capable systems circulate beyond their original developers?