The United States and China will hold talks on artificial intelligence in September, with Treasury Secretary Scott Bessent leading the US side. The discussions build on a safety framework the two countries agreed to in May.
What the September Talks Aim to Address
The upcoming meeting marks a continued effort to manage the risks around AI development. The May framework established a baseline for cooperation, but the specifics of the September agenda have not been made public. Both sides are expected to discuss how to prevent misuse of AI technologies while keeping economic competition in check.
Bessent's involvement signals that the US sees the talks as having a strong economic dimension. The Treasury Department has been increasingly focused on how AI could affect financial stability, trade, and national security.
The May Framework and Its Role
In May, the two countries outlined a set of safety principles for AI systems. That framework was a first step — it created a shared vocabulary around risks like bias, surveillance, and autonomous decision-making. The September talks are meant to move from principles to practical measures.
Neither side has released the full text of the May agreement, but officials have described it as covering transparency and testing requirements for advanced AI models. The new round of talks will test whether that framework can hold up under political pressure.
Why Treasury Leads the US Delegation
Bessent's role as the top US negotiator is unusual for AI talks, which often involve science or technology agencies. But the Treasury secretary has made AI a priority, arguing that the technology's economic impact is too big to leave to regulators alone. His team will likely push for guardrails that protect US financial markets and intellectual property while keeping channels open with Beijing.
China has not yet named its lead negotiator. The talks are expected to be held in person, though the location has not been announced.
The September meeting is not a formal treaty negotiation. It's a working-level dialogue meant to keep the two largest AI powers from drifting into open conflict over the technology. Whether that's enough to slow an arms race in AI development remains an open question — one the talks will try to answer.
