The United States and China have agreed to create a direct communication channel for serious incidents involving artificial intelligence, establishing a narrow guardrail between the world’s two largest AI powers after President Donald Trump’s September summit with Chinese President Xi Jinping.
A White House fact sheet published September 25 says the two countries also formed a bilateral dialogue to exchange views on AI’s risks and benefits. The next exchange is expected by November 2026.
The operating details remain open
The fact sheet does not define which events qualify for the incident channel, identify the agencies responsible for it or specify response times. Earlier U.S. discussions described a notification mechanism for AI incidents rising to the level of national security. Examples raised by experts included AI-enabled cyberattacks, biological misuse, major model failures and loss of human control.
That makes the agreement closer to a crisis-communication framework than a treaty restricting AI development. It creates a route for notification and clarification when an incident could be misread as hostile action, but the channel’s value will depend on whether both governments use it promptly during a real emergency.
Cooperation inside a larger rivalry
The agreement arrives while the two countries remain competitors in advanced models, chips and deployment. China’s official account of the summit emphasized continued dialogue, preventing misuse and keeping humans in control of AI systems. The White House document uses the administration’s preferred term “super intelligence” for the covered technologies.
Neither side announced a pause in model development or a shared technical standard. The immediate achievement is smaller and more practical: a standing diplomatic path intended to reduce opacity when an AI event has national-security consequences.
