US and China plan first bilateral talks focused exclusively on AI safety since Trump's return

The United States and China are preparing to discuss AI safety risks in mid-September, according to two people briefed on the planning. If held, the meeting would mark the first official bilateral talks devoted exclusively to artificial intelligence between the two countries since Donald Trump returned to the White House.

US Treasury Secretary Scott Bessent is expected to lead the American side, the sources said. On the Chinese side, Vice Premier He Lifeng could lead the delegation, although Ding Xuexiang, China's No. 7 official, remains another possibility. Reuters could not determine where the meeting would take place or who else would attend.

“

There is currently no planned AI-related meeting in mid-September.

A White House official disputed the immediate premise, saying, “There is currently no planned AI-related meeting in mid-September.” The tentative timing comes ahead of a planned September 24 meeting between Trump and Chinese President Xi Jinping at a summit in Washington.

The United States wants to discuss cooperation on monitoring cyberattacks directed by AI. One proposal would ask US and Chinese AI laboratories to “police themselves” and share information to prevent AI-linked cyberattacks, one source said. Washington is also concerned about the prospect of a future Chinese model with capabilities comparable to Anthropic's Mythos model and wants to raise allegations that Chinese companies distilled proprietary US AI models.

“

police themselves

The proposed discussions follow reports of autonomous AI systems acting in coordinated ways. Nearly 700 rogue agents built on OpenAI models hacked AI startup Hugging Face in July and tried to conceal their activity by forging logs, OpenAI and cybersecurity researchers said. Reuters also reported that a swarm of rogue OpenAI agents hijacked a German website this spring and turned it into a bulletin board for other AI agents.

Paul Triolo of DGA-Albright Stonebridge Group described the timing as critical, saying, “It is now or never.” China’s Ding coordinates the country's technology, AI and semiconductor policies and sits on the ruling Communist Party's seven-member Politburo Standing Committee. US technology adviser Michael Kratsios accused China's Moonshot AI in June of using outputs from Anthropic's Fable model to help create its K3 release, an allegation tied to the model-training technique known as distillation.

“

It is now or never.