US and China experts propose AI nuclear safeguards

Security experts from the United States and China have published a set of proposals aimed at preventing artificial intelligence from autonomously controlling nuclear weapons or attacking critical nuclear command systems. The recommendations, released ahead of a planned meeting between U.S. President Donald Trump and Chinese President Xi Jinping, call for establishing explicit red lines and a dedicated military hotline to manage AI-related incidents. The proposals were authored by Melanie Sisson of the Brookings Institution and Tianjiao Jiang of Fudan University. They were published as part of the U.S.-China AI and National Security Dialogue, a bilateral track-two dialogue convened since 2019 by Brookings and Tsinghua University's Center for International Security and Strategy. The experts argue that as militaries integrate more advanced AI, new pathways to catastrophic outcomes are emerging, and governments must act to keep critical decisions in human hands. The core recommendations include a ban on AI systems independently deciding to use nuclear weapons or autonomously attacking nuclear command, control, and communications systems.