FAR.AI is a research organization focused on frontier alignment research, working to improve the safety and security of artificial intelligence systems. Its technical work spans model evaluation, interpretability, robustness, and value alignment. The organization has produced over 40 influential academic papers in these domains.
The organization operates through several programs. FAR.Research pursues a diverse portfolio of high-potential research agendas. FAR.Labs is a coworking space in Berkeley designed to foster collaboration among researchers, policymakers, and industry leaders. FAR.Futures runs global events and policy initiatives to bring these same groups together on an international scale.
FAR.AI is based in Berkeley with a global reach, operating at the intersection of AI safety research and policy.
