US-China AI Hotline Proposal Leaves Critical Questions Unanswered

Washington has proposed warning Beijing about AI incidents that could threaten national security, but no agreement exists on what counts as an emergency or who must pick up the phone.

The United States has proposed a direct notification channel with China for artificial-intelligence incidents that could threaten national security. But Beijing has not publicly accepted the mechanism, and neither government has said what would trigger an alert.¹²

Treasury Secretary Scott Bessent announced the proposal after talks with Chinese Vice-Premier He Lifeng in New York on September 20.
The mechanism is still a US proposal, not an operational agreement or treaty.
The stated scope is limited to AI incidents with national-security implications; military systems, cyberattacks, critical infrastructure and model failures have not been formally included.
Earlier US-China discussions focused on monitoring AI-directed cyberattacks and sharing information between laboratories.
Separate US and Chinese experts have proposed stronger safeguards, including human control over consequential cyberattacks and a military hotline for autonomous-AI incidents.³

Bessent framed the idea as a way to make the world’s two leading AI powers less opaque to each other. Semafor reported that officials also discussed creating a formal bilateral AI dialogue and meeting again, but the public accounts contain no timetable, reporting procedure, named contact points or enforcement mechanism.¹

That gap matters. An AI system interfering with nuclear command networks, attacking critical infrastructure or launching a cyber operation could be mistaken for a deliberate act of war. Researchers and former officials have warned that automated systems may move faster than human crisis managers, while existing US-China military communication channels have not always worked during emergencies.³

For now, the proposal gives Washington a diplomatic channel to pursue, not a safety net. The next test is whether Trump and Xi turn Bessent’s announcement into a defined agreement — and whether either side is willing to report an incident that could expose a failure in its own military, intelligence or technology systems.

US-China AI Hotline Proposal Leaves Critical Questions Unanswered

Footnotes