US and Chinese Security Experts Propose AI Nuclear Control Rules
Brookings and Fudan University researchers urge the US and China to agree that only humans, not AI, can decide to launch cyberattacks on nuclear systems.
Why it matters
The proposal addresses a regulatory gap in AI governance between the world's two largest military powers, with direct implications for nuclear security protocols and future AI arms control negotiations. If adopted, the rules could shape international norms around military AI development and deployment.
Key facts
Two U.S. and Chinese security experts proposed rules barring AI from independently launching cyberattacks against nuclear command systems or deciding to use nuclear weapons.
Reported by Decrypt
A report by the Brookings Institution and Fudan University urges the U.S. and China to extend their commitment to human control of nuclear weapons to other military uses of AI, including cyberattacks on nuclear command systems.
Reported by Decrypt
The researchers wrote that Presidents Donald Trump and Xi Jinping should agree that only humans, not AI, decide to initiate cyberattacks against NC3 systems or critical infrastructure.
Reported by Decrypt
The authors called for a military AI hotline between the U.S. and China.
Reported by Decrypt
The authors called for a shared definition of human control over military AI.
Reported by Decrypt
NC3 systems help governments detect threats, decide whether to use nuclear weapons, and transmit launch orders.
Reported by Decrypt
Failures at any stage of NC3 systems could disrupt communications, misread an attack, encourage a first strike, or trigger an accidental or unauthorized launch.
Reported by Decrypt
AI's speed and uncertain behavior could leave leaders with little time to stop an escalating conflict.
Reported by Decrypt
What happened
Two security experts from the United States and China proposed red lines that would bar artificial intelligence from independently launching cyberattacks against nuclear command systems or deciding to use nuclear weapons, Decrypt reported. A report by the Brookings Institution and Shanghai-based Fudan University urges the two governments to extend their existing commitment to human control over nuclear weapons to cover other military uses of AI. The researchers wrote that Presidents Donald Trump and Xi Jinping could lead by agreeing that only humans should decide to initiate cyberattacks against each other's nuclear command, control, and communications systems (NC3) or critical infrastructure. The report also calls for establishing a military AI hotline and developing a shared definition of human control. NC3 systems help governments detect threats, decide whether to use nuclear weapons, and transmit launch orders. The authors warned that AI's speed and uncertain behavior could leave leaders with little time to stop an escalating conflict.
How the story developed
- First report
First report by Decrypt
Security Experts Want the US and China to Promise Never to Let AI Control Nukes
- Update
Brookings and Fudan University publish joint report on AI nuclear control rules
The report proposes red lines barring AI from independent nuclear cyberattacks and calls for a military AI hotline and shared definition of human control.
Affected entities
Original sources
| Publisher | Report | Role | Published |
|---|---|---|---|
| Decrypt Crypto media | Security Experts Want the US and China to Promise Never to Let AI Control Nukes | Original |