
US and Chinese Experts Urge Shared Rules to Keep AI Out of Nuclear Launch Decisions
Experts from the Brookings Institution and Fudan University have proposed shared US-China rules barring AI systems from autonomously launching nuclear weapons or attacking nuclear command systems. The recommendations, published ahead of a September 24 Trump-Xi meeting, include a hotline for AI incidents modeled on Cold War direct lines. They build on a November 2024 Biden-Xi understanding, though binding mechanisms remain absent and skeptics question whether crisis communication channels would be answered.
