US and China Plan Formal September AI Safety Talks Under President Donald Trump

Treasury Secretary Scott Bessent said, "The reason we are able to have wholesome discussions with the Chinese on AI is because we are in the lead," highlighting the U.S. position as a driver of these talks and his role in laying groundwork for an AI safety protocol in Beijing.
China’s foreign ministry confirmed in May that the two nations had agreed to establish intergovernmental AI talks, signaling official buy-in beyond informal exchanges.
The September talks would unfold amid ongoing Track II diplomacy and related discussions at the UN AI conference in Geneva, where U.S. and Chinese researchers addressed risks like AI-enabled cyberattacks and escalation concerns.
Regulatory dynamics around AI are intensifying: Axios and the Financial Times have reported moves in Washington to ban or restrict Chinese advanced AI models and to tighten export controls on AI and semiconductors, with Moonshot’s Kimi K3 open-source model fueling Silicon Valley anxiety.
The United States and China are planning their first official AI talks under President Donald Trump, set for September, according to Quartz and NDTV Profit. Treasury Secretary Scott Bessent will lead the American side, making him the top U.S. official to engage China directly on artificial intelligence risks.
The talks are expected to happen before Chinese President Xi Jinping visits the United States on September 24. Exact dates and locations are not yet set, but both governments have confirmed the talks are coming, Quartz reported.
The September talks trace back to a May summit between Trump and Xi. After that meeting, China's foreign ministry confirmed both nations agreed to set up formal, government-to-government AI discussions. That was a significant step — moving beyond informal backchannels into official diplomacy, Quartz reported.
Bessent has been laying the groundwork. He traveled to Beijing to build the foundation for what he described as an "AI safety protocol." He said, "The reason we are able to have wholesome discussions with the Chinese on AI is because we are in the lead," signaling that Washington sees itself as driving the agenda, according to Newsy Today.
The agenda is still being finalized, but officials say several key topics are on the table. These include AI safety and governance, export controls on chips and software, military uses of AI, and the risk of cyberattacks powered by AI systems, according to NDTV Profit.
One major concern is the spread of powerful AI to non-state actors — groups like terrorist organizations that could misuse the technology. U.S. and Chinese researchers have already begun addressing these risks at the United Nations AI conference in Geneva, including fears about AI-enabled escalation in conflicts, Quartz reported.
While diplomats plan talks, tensions are rising in the tech world. Washington is weighing a ban on Chinese advanced AI models and tighter export controls on AI and semiconductors, according to reporting by Seeking Alpha. The moves reflect growing unease about China's rapid AI progress.
One flashpoint is Moonshot's Kimi K3, a Chinese open-source AI model that has rattled Silicon Valley. Open-source models are AI systems anyone can download and modify freely. That makes them hard to control — and easy to spread. U.S. regulators are now weighing how to respond, Newsy Today reported.
The September talks would not solve the broader U.S.-China tech rivalry. But they do mark a shift — from competition with no guardrails to competition with at least some formal communication. Both sides appear to want rules around the most dangerous AI risks, even if they disagree on almost everything else.
Track II diplomacy — informal exchanges between researchers and experts — has already been happening, NDTV Profit noted. The September talks would formalize that process. With Xi expected in the United States on September 24, the pressure is on to show progress before leaders meet face to face.
Publishers
33
Articles
41
Reach
74