Rapidly advancing AI has prompted a wide spectrum of views about AI safety, ranging from existential fears to skeptical dismissals of “doomers.” Meanwhile, few are tracking the actual policy work that is starting to address these concerns. As the first day of a set of discussions about AI and society, the Hoover Institution is hosting a symposium on AI safety and catastrophic risk. Our speakers are those who are actually working on policy initiatives, conducting initial risk assessments, and developing methods to evaluate frontier models. We will discuss and illustrate the state of global work and the way leading companies are comprehending the policy challenge. This half-day forum will help inform participants as they navigate the risks and analyze how to sustain fruitful innovation and openness.