To agree to international AI red lines, we need to build the technology that makes it possible to adhere to them. In this episode, Tristan sits down with two experts in this field to discuss the kinds of verification technology we need for AI, the challenges of building it, and the world it could unlock if we do.