US and Chinese Security Experts Propose Nuclear-Style Safeguards for Military AI
Security experts from the US and China have proposed a set of safeguards for military AI modelled on the rules that grew up around nuclear weapons, including a ban on letting AI decide whether to launch them. (Reuters)
Key Recommendations:
- No Autonomy for AI in Nuclear Decisions: AI should not be allowed to autonomously decide to use nuclear weapons.
- Protection Against Cyberattacks: Humans should maintain control over any consequential cyberattack on a rival's nuclear infrastructure.
- Dedicated Military Hotline: Establish a dedicated US-China military hotline for AI incidents.
- Definining "Meaningful Human Control": Agree on what "meaningful human control" of a weapons system actually means, as this term is widely used in arms-control debates but lacks a shared definition.
Background:
These proposals come from the US-China AI and National Security Dialogue, run since 2019 by the Brookings Institution and Tsinghua University’s Center for International Security and Strategy. The ideas build upon an agreement reached in November 2024 between then-President Joe Biden and Chinese President Xi Jinping, affirming that humans, not AI, should control decisions to use nuclear weapons.
Timing and Significance:
The proposals are timed to coincide with a planned meeting between Donald Trump and Xi Jinping on September 24th and a fresh round of official AI talks where the US intends to raise the issue of AI-directed cyberattacks. They also come amid growing political debate over AI development, with US lab chiefs calling for a slowdown and China's commerce ministry dismissing that plan as technological hegemony.
Aim:
The Brookings-Tsinghua group aims to keep the most dangerous military decisions in human hands.