BEIJING — China's Commerce Ministry confirmed on Thursday that senior Chinese and American trade negotiators have held their first talks focused on artificial intelligence, a milestone in the two countries' AI diplomacy that came just hours before President Donald Trump and Chinese President Xi Jinping were scheduled to begin talks in Washington.
The confirmation, delivered by ministry spokesperson He Yadong at a Thursday briefing, marks the first time either government has publicly acknowledged that AI-specific discussions between the two delegations have taken place. According to CNBC, He said the two sides also discussed plans for reducing tariffs and for extending trade arrangements agreed last fall. For more context on this story, see our ongoing coverage of the latest AI developments.
What happened at the first AI talks
He was referring to meetings between Chinese Vice Premier He Lifeng and US Treasury Secretary Scott Bessent in New York in the run-up to this week's summit between the two presidents. The ministry described the exchanges as "constructive, candid talks" and said the two sides "reached multiple points of consensus," without disclosing what those points were.
According to Yahoo News' account of the negotiations, which cites NBC News reporting, the September dialogue had been planned since the May Trump-Xi summit in Beijing, where both governments agreed to establish intergovernmental AI talks. Before this week, no AI discussions between the two countries had been publicly confirmed.
A trade truce extended to January
The AI milestone landed alongside a broader economic de-escalation. Bessent told Fox News on Wednesday that the two countries had agreed to extend their bilateral trade truce through January 10, according to NBC News.
"We will extend what we call the 'Busan Agreement' — the economic détente between the two countries that was scheduled to end on Nov. 10 — that is going to be extended until Jan. 10," Bessent said, per NBC News. He added that January could bring either a broader economic agreement or simply a continuation of the existing arrangement.
The truce, struck last fall, kept tariffs on Chinese goods at around 20 percent and included a Chinese commitment to resume shipments of critical minerals — rare earths that serve as essential inputs for semiconductors, consumer electronics, and defense products. A senior US official said ahead of the summit that China's delivery of those minerals "has not been up to par," according to Yahoo News.
Bessent also suggested the Chinese side might have additional announcements to make over the next several days, among them possible commitments to expand purchases of US farm goods and to open parts of its financial services sector, according to NBC News.
An AI risk alert mechanism in the making
Earlier in the week, Bessent said the two sides had discussed establishing a formal AI dialogue and a mechanism to alert each other about AI-related risks. Thursday's confirmation indicates that conversation is now formally underway.
The diplomatic outreach comes as both governments weigh how to address the risks posed by rapidly advancing AI systems. According to CNBC, recent incidents involving AI models have raised fears that increasingly autonomous systems could make attacks faster and harder to contain — a concern that has pushed AI from a commercial file to a core national-security topic between Washington and Beijing.
Trump: "Our guardrail is the DOJ"
Trump flagged AI as central to the summit in a Truth Social post early Thursday. "A big day with President Xi of China. Super Intelligence (SI) will be a big topic of discussion, but I want to leave it exactly where it is. That is China's position also. Our guardrail is the DOJ!" the president wrote.
The post is consistent with Trump's public posture on AI regulation. Speaking at the UN General Assembly on Tuesday, he said: "We're going to encourage it, not rein it in," according to CNBC.
Why the first US-China AI talks matter
The establishment of an AI channel between the two countries carries weight for at least three reasons.
First, guardrails: an alert mechanism for AI risks echoes the hotlines of nuclear-era arms control — a standing way to flag incidents quickly before they escalate, in a domain where autonomous systems could act faster than diplomats can.
Second, leverage: the AI compute race runs on rare earths and advanced semiconductors, the same supply chains at the center of the truce. Extending the truce keeps that trade flowing at least through January, and ties AI cooperation to the minerals that make it physically possible.
Third, divergence: Washington's stated position is to encourage AI development with minimal new regulation, while Beijing runs a state-directed model. A standing dialogue creates at least a shared floor for managing incidents between two rapidly divergent regulatory systems.
What remains unclear is the substance. The "multiple points of consensus" were not disclosed, and the proposed alert mechanism's scope — what counts as an AI risk worth flagging, and who on either side receives the alert — has not been defined publicly.
Observers will now watch whether the Washington summit produces a named framework for the AI dialogue, and whether either side offers any detail on the incidents such a mechanism would be designed to catch.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
