Analysts Urge US-China AI Nuclear Hotline
Experts call for a dedicated channel to reduce the risk of artificial intelligence triggering a nuclear attack between the two powers.
Updated on Nov. 12, 2026 in Artificial Intelligence

Live Poll
Should the US and China agree to keep AI out of nuclear weapons decisions?
On Wednesday, US Secretary of State Marco Rubio and Chinese counterpart Wang Yi agreed that leader-level conversations would help advance peace and denuclearisation. Analysts speaking at a Shenzhen forum on Saturday called for a dedicated hotline to address the risk of AI-linked nuclear escalation.
Why it matters
An intercontinental ballistic missile would take around 30 minutes to travel between China and the US, leaving little time for human response to a false alarm. Commentators argue that the same concerns that shaped Cold War crisis management now apply to artificial intelligence in nuclear command systems.
The only quantified technical reference is the roughly 30-minute travel time of an intercontinental ballistic missile between China and the US, which sets the window in which AI-assisted decisions would have to be made. Whether AI systems would be connected to nuclear command structures has not been disclosed by either government.
The players
Marco Rubio
US Secretary of State who held a phone call with Chinese counterpart Wang Yi on arms control and denuclearisation.
Wang Yi
China's foreign minister who agreed with Marco Rubio that leader-level conversations would help advance peace and denuclearisation.
Donald Trump
The current President of the United States, who renewed his call to discuss nuclear arms controls with Xi Jinping.
Xi Jinping
President of China, expected to meet Donald Trump at the planned APEC summit in Shenzhen.
The details
Zhou Bo, a retired senior colonel in the People's Liberation Army, told a Shenzhen forum on Saturday that military use of AI is already a reality and proposed monthly government discussions through a dedicated hotline. He said Beijing and Washington could discuss ways to prevent AI from controlling nuclear command systems, while a PLA Daily commentary on Thursday called for a governance framework with broad participation and consensus.
Timeline
1983: Soviet officer Stanislav Petrov judged a satellite warning of incoming US missiles to be incorrect and did not escalate.
Wednesday: Rubio and Wang Yi held a phone call and agreed leader-level conversations would help advance denuclearisation.
Saturday: Zhou Bo spoke at the Shenzhen forum on military AI.
Thursday: The PLA Daily published its commentary calling for AI governance.
Upcoming: Donald Trump is expected to meet Xi Jinping at the APEC summit in Shenzhen.
The Tech Race
The 1983 incident, in which Soviet officer Stanislav Petrov dismissed a false US missile warning, is cited as proof that human judgment has averted nuclear catastrophe. Analysts now argue that AI-assisted command systems would shrink the time available for such judgment well below the roughly 30 minutes an intercontinental missile needs to cross between China and the US.
For ordinary readers, the debate matters most as a question of whether AI will be allowed to touch the systems that decide nuclear launches. Any future hotline or agreement would shape how much of that decision stays with people rather than software.
The takeaway
Keeping humans in the loop on nuclear decisions is the core lesson, and direct government-to-government channels are one of the few tools proposed to make that possible.
Further reading
Explore more reporting in our Artificial Intelligence section.
Live Poll
Should the US and China agree to keep AI out of nuclear weapons decisions?







