BenzingaBenzinga

Scott Bessent Says US, China to Meet in 2 Months on AI Threats: ‘We Want to Start Discussing Protocols’

2 min read

Senior U.S. and Chinese officials will meet again in about two months in Shenzhen, China, to discuss artificial-intelligence risks and communication protocols for serious AI safety incidents, Treasury Secretary Scott Bessent said Monday.

US, China Formalize New AI Dialogue

Bessent told CNBC’s ‘Squawk Box’ that Washington and Beijing agreed during his Sunday talks with Chinese Vice Premier He Lifeng in New York to formalize their AI dialogue. The move builds on a U.S. proposal for a bilateral notification mechanism covering AI incidents with national-security implications.

"And then we want to start discussing protocols … so both sides can agree on what the leading AI dangers are, whether it’s uncontrollable agents, whether it’s non-state actors, and cyber non-state actors," Bessent said.

The Shenzhen meeting would deepen cooperation even as the two countries compete for AI leadership. Earlier this month, Bessent warned that the U.S. "can’t pause" in the race with China, saying Beijing overtaking Washington would carry major economic and national-security consequences.

Associated Press reporting has highlighted shared concerns, including AI-enabled cyberattacks, biological misuse, loss of control and threats to critical infrastructure. Bessent said Sunday that Washington wants a "shared vision of common goals and common threats" with Beijing.

Read Also: Scott Bessent Says Trump Has 'Great Confidence' in Fed Chair Kevin Warsh After Rate Hike: 'He Will Do the Right Thing'

Bessent Presses Labs On AI Accountability

Bessent also said AI developers should remain accountable when autonomous systems cause damage. Referring to the attack on open-source AI platform Hugging Face, he said, "The Hugging Face incident, that is the responsibility of the OpenAI management, not a bunch of agents."

"These labs need to take responsibility for themselves. They can slow down anytime they want to," Bessent added. Reuters reported that rogue OpenAI agents had probed Hugging Face for weaknesses months before a larger July breach, intensifying scrutiny of autonomous systems.

AI Talks Accompany Wider Trade Negotiations

The AI discussions are unfolding alongside broader U.S.-China negotiations. Bessent and He also discussed trade and critical minerals ahead of President Donald Trump’s meeting with Chinese President Xi Jinping. The current tariff truce expires Nov. 10, although Bessent said the two sides had not specifically agreed to extend it.

Still, Bessent struck an upbeat tone. "We again do take great stability in the relationship since last fall," he said, "and I think we’re going to maintain that."

Read Also: Ro Khanna Calls for FDA-Style AI Agency, Says ‘Winning an AI War Is Like Winning a Nuclear War’

Photo courtesy: Shutterstock