Scott Bessent says China won’t disclose AI incidents, pushes for bilateral notification system

US Treasury Secretary Scott Bessent has put a sharp point on something Washington has long suspected: China would not voluntarily disclose if it experienced significant AI incidents. The statement, made during discussions with Chinese Vice Premier He Lifeng in New York on September 20, 2026, frames the core problem that a proposed bilateral AI notification mechanism is meant to solve.

The talks concluded with both sides exploring a system where the US and China would notify each other about national security-level AI incidents.

What Bessent is actually proposing

The notification mechanism would create a formalized channel for the world’s two largest AI developers to flag serious incidents to each other. Bessent framed the goal as building a “shared vision for common goals and threats” between the US and China on artificial intelligence.

This isn’t a proposal that materialized out of thin air. It builds on earlier discussions between President Trump and President Xi from May 2026, where establishing a broader “US-China AI dialogue” was first floated.

One notable carve-out: export controls on advanced AI chips are explicitly excluded from the proposed dialogue.

The bigger picture: distillation and the technology race

Bessent’s push for transparency arrives against a backdrop of escalating US concerns about Chinese AI development practices. A central worry in Washington involves what’s called AI model “distillation,” where Chinese firms allegedly leverage outputs from advanced US AI systems to accelerate training of their own models.

Bessent himself warned earlier in mid-2026 that China’s AI advancements represent America’s “biggest risk” in technology.

What this means for markets and policy

The exclusion of chip export controls from the dialogue creates a split-screen dynamic for semiconductor companies. Those firms remain caught between enormous demand from AI development and tightening government restrictions on who they can sell to. The fact that Washington chose to keep chip policy separate from the AI safety conversation suggests the administration views export controls as a leverage tool it doesn’t want to negotiate away at the safety table.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.