Policy

US and China Agree to Create AI Safety Dialogue Channel

Treasury Secretary Bessent announces notification mechanism for AI incidents as world's top two AI powers seek transparency amid competition.

Omega Editorial· September 21, 2026· 3 min read

US and China Move Toward AI Safety Communication

The United States and China have agreed to establish a formal communication channel dedicated to artificial intelligence concerns, US Treasury Secretary Scott Bessent announced following economic talks in New York on Sunday.

The proposed "US-China AI dialogue" would create a notification mechanism between the world's two leading AI powers, allowing them to alert each other about potential incidents that reach national security significance. Bessent met with Chinese Vice Premier He Lifeng to lay groundwork ahead of a Thursday White House summit between President Donald Trump and Chinese President Xi Jinping.

"We want a shared vision of common goals and common threats," Bessent told reporters. "We think that just like any cross-border activity, that moving from opaque to more transparency between the No. 1 and the No. 2 AI powers in the world is very important."

Why It Matters

This represents a rare area of potential cooperation between Washington and Beijing despite their broader technological rivalry. As AI systems grow more capable, both nations face shared existential risks that transcend competitive interests. A notification system could prevent miscalculation or provide early warning of AI-related incidents that might otherwise escalate into international crises. The mechanism also acknowledges that AI safety challenges require coordination even between adversaries.

Details of the Proposed Framework

The notification system would focus on situations that rise to national security levels involving artificial intelligence, though specific thresholds and protocols remain to be defined. Both sides committed to additional meetings to operationalize the dialogue.

The talks also addressed establishing a "Board of Trade" that Trump and Xi had previously discussed during their May meeting in Beijing, suggesting broader economic engagement alongside the AI-specific channel.

Chinese state media characterized the Sunday discussions as "candid, in-depth, and constructive," according to Xinhua news agency. The official readout confirmed that officials held dialogue on AI-related issues but provided no additional specifics beyond noting discussions on bilateral trade and investment matters.

Balancing Competition and Safety

The initiative emerges against a backdrop of intense technological competition. President Trump has resisted domestic calls to slow AI development, arguing that restraint would allow China to close the gap with US companies. The proposed dialogue suggests both nations recognize certain risks require collaboration despite their race for AI supremacy.

Bessent emphasized the importance of transparency between the two dominant AI powers, framing the communication channel as essential infrastructure for managing cross-border AI activity in an era of rapid technological advancement.

The agreement sets the stage for Thursday's presidential summit, where the AI dialogue framework may receive formal endorsement at the highest levels of government.

These details were first reported by Euronews.

#us-china relations#ai safety#ai governance#national security#international cooperation#ai policy

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Policy

Policy· 3 min read

U.S. and China Seek AI Safety Dialogue Amid Tech Rivalry

Treasury officials propose notification mechanism for AI incidents as Trump and Xi prepare to meet, though export controls remain firmly in place.

Via AI Watch · Sep 21, 2026
Policy· 3 min read

US Cloud Computing Restrictions Could Choke China's AI Sector

Washington's potential expansion of tech blockades to cloud networks would create the most severe bottleneck yet for Chinese AI developers already cut off from advanced chips.

Via AI Watch · Sep 21, 2026
Policy· 3 min read

OpenAI Execs Called AI 'Existential Threat' to Publishers

Unsealed court documents reveal internal acknowledgments that chatbots substitute for news sites and circumvent paywalls.

Via AI Watch · Sep 21, 2026