U.S. and China agree to create a channel for serious AI incidents

The United States and China have agreed to open a formal dialogue on what the White House calls “super intelligence” and create a direct channel for serious incidents involving advanced AI. The dialogue will cover the technology’s risks and benefits, with another exchange planned by November. As the two countries compete to build more capable systems, the channel could give their governments a way to clarify a dangerous event before it is misunderstood or grows into a wider crisis. It is a limited safeguard, not a shared safety regime or a deal to slow the AI race: neither side has said what qualifies as an incident, who would use the channel, or what information either government would have to disclose.