A key element of AI safety? Democracy | Matthew Botvinick

Matthew Botvinick argues that constitutional democracy is an indispensable prerequisite for AI safety: even if AI could no longer escape human control, launch cyberattacks, or manufacture bioweapons, under an authoritarian regime it would still become a tool for surveillance, censorship, and political repression. He calls for amending the U.S. National Emergencies Act and the International Emergency Economic Powers Act, closing the "data broker" loophole in intelligence work, and equipping Congress and the courts with equally powerful AI tools to maintain checks and balances.