A US proposal has become a joint commitment. Following the meeting between Donald Trump and Xi Jinping, the White House said on September 25 and China's foreign ministry said on September 26, 2026 that the two countries had agreed to establish a direct communication channel for AI incidents.
Both statements also describe a dialogue on the technology's risks and benefits, with the next exchange planned for November 2026. The US calls it a Super Intelligence dialogue and refers to “SI incidents”; the Chinese statement says “AI incidents”. Neither explains whether that difference will affect what the channel covers.
Days earlier, US Treasury Secretary Scott Bessent had proposed notifications for AI events reaching the national-security level. At that point, China's acceptance was not public. The new statements from both governments close that evidentiary gap: they confirm an agreement to establish a channel. They do not establish that it is operating yet.
Neither government has published a reporting threshold, designated contacts, a notification deadline or rules for what information would be exchanged. There is no announced start date. Without those details, the practical effect of the commitment cannot yet be assessed.
This is a Risk Signal: a policy response to possible AI security events, not evidence of a new incident in which an AI system acted on its own. It does not change the incident count or timeline.