US and China experts propose nuclear-style controls on military AI

September 17, 2026 at 04:19 PM2 min

This digest was compiled by AI from multiple sources — links to the originals are below.

US and Chinese national security experts have proposed a package of measures to reduce military AI risks, modeled on nuclear arms control. The proposals include red lines barring AI from autonomous nuclear weapons decisions and a dedicated hotline for AI incidents. The report was released ahead of a planned Trump-Xi meeting on September 24 in Washington.

Key Facts

The report was authored by Melanie Sisson of the Brookings Institution and Tianjiao Jiang of Fudan University.
The experts propose 'red lines' that would prohibit AI from autonomously deciding to use nuclear weapons or attack nuclear command systems.
Tianjiao Jiang proposed a dedicated military hotline between the US and China for reporting incidents involving autonomous systems.
Neither Washington nor Beijing has officially endorsed the initiative.
China's disarmament ambassador Shen Jian said in June at a UN meeting in Geneva that military AI could affect strategic stability and increase the risk of miscalculation and escalation.

Proposed Safeguards

The experts' key concern is a scenario in which an AI system interferes with nuclear command or launches a military cyber operation, leaving Washington or Beijing only minutes to determine whether it was a real attack. The proposed red lines would prohibit AI from autonomously deciding to use nuclear weapons or attack nuclear command systems. Melanie Sisson argued that humans must retain the authority to sanction significant cyberattacks against nuclear command, control, and communications, as well as strategically important infrastructure. Tianjiao Jiang proposed a dedicated military hotline between the US and China for reporting incidents with autonomous systems, allowing one side to explain that unusual AI activity was accidental or under investigation before the other side treats it as deliberate aggression.

Institutional Context

China has already incorporated AI into its arms control system, placing it under the portfolio of a Foreign Ministry department that also handles nuclear weapons and nonproliferation. The United States has no single center of responsibility for AI, with the topic overseen by the National Security Council, the State Department, and the Pentagon. China's disarmament ambassador Shen Jian said in June at a UN meeting in Geneva that military AI could affect strategic stability and increase the risk of miscalculation and escalation, and that weapons must remain under human control. China's Minister of State Security Chen Yiwen made a similar point this week, calling AI a factor that fundamentally changes the nature of military confrontation.