US and China Launch AI Dialogue as Bessent Warns Against Illusions

The United States and China have established a new dialogue mechanism on artificial intelligence aimed at averting catastrophic failures, even as both governments race for dominance in the technology, Treasury Secretary Scott Bessent said Tuesday. The two countries should "harbor no illusions" about their differences as they cooperate, Bessent said in prepared remarks to the National Committee on U.S.-China Relations in New York.

The stakes are considerable. Washington and Beijing are the two leading developers of frontier AI systems, and neither has committed to slowing down. What they have built instead is a communications channel designed to keep competition from turning into miscalculation.

What does the new mechanism cover?

Bessent described the dialogue as focused on what he called "super intelligence," or SI. He said the mechanism aims to guard against "shared risks" and to avert any "misunderstanding between our two systems."

Under the arrangement, both sides will exchange "best practices" for testing advanced models before release. Bessent said the mechanism is built around "keeping human judgement firmly in command of the technology." He also said he expects both governments to alert each other to serious incidents involving SI models — a hotline-style arrangement modeled on Cold War-era crisis communication.

Bessent stopped short of any commitment to limit advancement of the technology. That omission reflects the political reality in both capitals: Beijing and Washington remain locked in an intense race to achieve global dominance in AI, and neither side will accept constraints the other refuses to match.

Where does Trump stand on guardrails?

The Trump administration has largely resisted broadening restrictions on AI development over concerns about ceding a competitive edge to China. The president has more recently signaled some openness to regulatory safeguards.

"I don't mind guardrails," Trump said in a CBS interview released Sunday. He added that he did not want regulations to put the U.S. in "a deep second position" behind China, which he said faced no such restrictions.

That framing — safety measures acceptable, unilateral handicap not — defines the administration's bargaining position as the dialogue proceeds.

How has China responded?

China's foreign ministry in September dismissed calls to slow down frontier models. It said fearmongering and confrontation would only disrupt global AI governance.

Huiyao Wang, founder of the Center for China and Globalization, a think tank that closely aligns with Beijing's thinking, said last month that the U.S.-China talks should "confront AI risks and manage them together."

Xie Feng, China's ambassador to the U.S., addressed the same New York audience. He said efforts by both countries were "indispensable" to addressing issues from global flashpoints to "improving artificial intelligence governance." Cooperation "may not resolve every challenge," Xie said, "but none can be resolved without it."

Why does this matter now?

During Xi's September visit to the United States, Washington and Beijing agreed to launch an AI dialogue with a communications channel to help defuse serious incidents. Tuesday's speech is the most detailed public account yet of what that channel will do in practice.

The two leaders are expected to meet for the third time next week, with Bessent joining, on the sidelines of the Asia-Pacific Economic Cooperation Leaders' Meeting from Nov. 18 to 19.

Bessent used the speech to set expectations for a relationship that will mix cooperation and rivalry indefinitely. As the technology advances, he said, the two nations will need to demonstrate how they can "reconcile the demands of competition with the responsibilities of leadership."

"Our interests will not always coincide, and our approach to every challenge likewise will not always converge," Bessent said.

What is the competitive backdrop?

The dialogue sits atop an unresolved technology conflict.

Washington briefly suspended access by foreign nationals to Anthropic's Mythos and Fable 5 models in June before restoring it weeks later. The U.S. government has also restricted exports of the most advanced chips to Beijing. And it has accused Chinese developers of copying American models through what it calls industrial-scale distillation — extracting the capabilities of a leading system into a cheaper replica.

Those measures remain in force. The new dialogue covers risk management, not export controls or model access.

Does the Jensen Huang scandal hang over the summit?

The New York event unfolded as controversy around Nvidia Corp CEO Jensen Huang's affair with an alleged Chinese spy continues to make headlines.

Huang admitted to an extramarital affair with a Chinese businesswoman whom U.S. intelligence officials suspect of espionage. The admission has placed the billionaire tech executive at the center of a major corporate and national security scandal.

Huang had joined Trump's delegation to Beijing in May and was invited to the state dinner in Washington in September. His presence at the upcoming summit has been thrown into doubt.

Other business leaders could take his place on the delegation. According to South China Morning Post, Tesla and SpaceX CEO Elon Musk and Mastercard's Michael Miebach could be on Trump's delegation next week.

What comes next?

The immediate test of the mechanism arrives within days: whether the Nov. 18-19 APEC meetings produce concrete incident-response protocols, or leave the dialogue as a statement of intent between two systems that, in Bessent's words, will not always converge.