US and China to set up 'communication channel' for AI incidents after summit

The United States and China will set up a "communication channel" for artificial intelligence incidents, the White House said on Friday, after President Donald Trump and President Xi Jinping concluded a summit that focused heavily on the rapid advance of the technology.

According to a White House fact sheet, the two countries will establish a "U.S.-China Super Intelligence (SI) Dialogue to exchange views on risks and benefits" of the technology. The leaders agreed to use the term "SI", short for super intelligence, rather than "AI" in their discussions, the fact sheet said. The next exchange is expected by November.

The two sides also agreed to create a "bilateral communication channel" for AI-related incidents. The White House did not detail how the channel would operate, who would staff it, or what would qualify as an incident.

Mr. Xi, speaking at the White House on Thursday, struck a measured tone, calling on the world's two largest economies to manage their rivalry and keep advanced AI under human control.

"We have both the capability and responsibility to develop and manage AI for good and ensure that the development of AI is always under human control," he said.

Mr. Trump had said before the summit that super intelligence would be a major topic of discussion. "Super Intelligence (SI) will be a big topic of discussion, but I want to leave it exactly where it is," he wrote on social media, adding that this was also China's position.

The idea of a dedicated line between two rivals is not new. During the Cold War, Washington and Moscow set up direct hotlines to reduce the risk of accidental escalation, and the U.S. and Chinese militaries have maintained communication links in recent years to prevent miscalculation. Officials and researchers have argued that as AI systems become more capable and more deeply embedded in critical infrastructure, a misread signal or an unanticipated failure in one country could quickly affect the other.

Both countries are competing to build the most advanced AI systems. Washington has restricted exports of advanced semiconductors and chipmaking equipment to China, while Beijing has pushed to develop its own models and hardware. At the same time, the two governments have held talks on the safety of frontier AI since 2023, when they agreed to open an intergovernmental dialogue on the risks of the technology.

The announcement did not define what would count as an "incident", nor did it set out joint safety standards or limits on the development of such systems. The fact sheet described the dialogue as a forum for exchanging views rather than a framework for binding commitments, and did not say whether the channel would involve technical experts, diplomats or military officials.

The agreement comes amid wider international efforts to set ground rules for AI. The United Nations General Assembly adopted a resolution in 2024 on the safe use of artificial intelligence, and several countries, including India, have called for greater international cooperation on AI safety.

For now, the practical effect of the agreement will depend on how the channel is used, and whether both sides treat it as a working tool for de-escalation rather than a symbolic gesture. The exchange due by November is likely to offer the first indication.