The article argues that focusing solely on detecting AI use in education misses the larger shift towards technical capabilities and institutional pipelines that transform research into products. It suggests a new approach is needed for higher education in an AI-transformed world.