Multiple efforts persist across institutions, but many function in silos. There is a growing need for a nodal mechanism to anticipate, mitigate and respond to AI-related risks, backed by uninterrupted government and stakeholder support and rapid access to emerging frontier models