US Lawmakers Back AI Safety Cooperation With China While Urging Tougher Chip Export Controls

Donald Trump -Xi Jinping

U.S. lawmakers from both political parties say Washington can cooperate with Beijing on establishing international artificial intelligence safety standards while maintaining strict restrictions on China’s access to advanced American semiconductor technology.

The comments come as President Donald Trump prepares to meet Chinese President Xi Jinping on Thursday on the sidelines of the Asia-Pacific Economic Cooperation (APEC) summit in Shenzhen, China. The meeting is expected to address a range of bilateral and global issues, including technology competition, trade and national security.

The debate over cooperation has gained renewed attention following discussions between China-based AI safety organization Concordia AI and several U.S. technology companies, including OpenAI, on international collaboration and the progress of AI safety initiatives.

Several lawmakers told reporters on Wednesday that Trump should use his meeting with Xi to raise concerns about China’s efforts to circumvent American semiconductor export restrictions. However, they stressed that preventing China’s access to advanced computing technology should not prevent cooperation on broader AI safety challenges.

“Our most powerful chips landing in the hands of China is a different conversation than us collaborating globally around AI safety standards,” said Representative Lori Trahan, a Massachusetts Democrat who introduced legislation requiring oversight of advanced AI models. “So I think we can do both of those things.”

The distinction reflects growing concerns in Washington over China’s expanding artificial intelligence capabilities and the potential national security implications of supplying its technology industry with advanced American chips.

The issue has attracted particular scrutiny following controversy involving Nvidia and former CEO Jensen Huang, renewing questions about the effectiveness of safeguards governing the company’s technology sales to China and the Trump administration’s enforcement of export controls.

Representative Sam Liccardo, a California Democrat representing Silicon Valley, argued that international AI safety cooperation should begin with technical discussions involving leading experts from both countries.

“We need to get those with expertise, top AI scientists on both sides of the ocean, at the same table to resolve the critical technical issues,” Liccardo said.

He suggested that technical specialists should first identify areas of agreement before political leaders become involved in establishing broader international arrangements.

Meanwhile, Republican lawmakers pushing legislation to restrict China’s access to advanced artificial intelligence computing infrastructure want Trump to take a tougher position during his discussions with Xi.

Senator Jim Banks, a Republican from Indiana and co-sponsor of the AI OVERWATCH Act, which seeks to impose export restrictions on advanced semiconductor technology to foreign adversaries, urged Trump to raise concerns surrounding the Nvidia controversy.

Banks said Washington should warn Beijing that “espionage activities must come to a stop, or we’re going to hold you accountable for it.”

Senator Pete Ricketts, a Republican from Nebraska who introduced the MATCH Act to tighten restrictions on exports of semiconductor manufacturing equipment, also supported discussions on AI safety between the two governments.

AI guardrails are “certainly worth” discussing between Trump and Xi, Ricketts said, adding that Washington and Beijing should begin conversations about establishing broad principles and safeguards governing the emerging technology.

Cooperation on artificial intelligence was also expected to feature during Xi’s September state visit to Washington. However, the two leaders failed to reach a formal agreement.

The main tangible outcome was a notification mechanism between U.S. Treasury Secretary Scott Bessent and his Chinese counterpart, allowing the officials to contact each other in the event of an artificial intelligence incident with potential national security consequences.

Some members of Congress, including Representative Greg Casar, a Texas Democrat, have advocated for expanding cooperation between the two countries as artificial intelligence becomes increasingly influential in economic, military and technological competition.

Despite concerns surrounding Nvidia and China’s access to advanced computing technology, several lawmakers believe Washington can strengthen export controls without abandoning international engagement on AI safety.

Representative Jay Obernolte, a California Republican, said continued dialogue with Beijing remained necessary even as Congress examines potential weaknesses in America’s semiconductor export-control system.

“There’s still a need to engage with China,” Obernolte said.

He also argued that Congress should investigate Nvidia’s activities and the Commerce Department’s role in enforcing restrictions on advanced semiconductor exports.

However, he emphasized that the outcome of those investigations should not undermine broader international cooperation.

“That might inform future decisions about export restrictions, but doesn’t alter the need for international engagement on safety standards,” he said.

The competing priorities highlight Washington’s increasingly complex approach to artificial intelligence: limiting China’s access to strategically sensitive American technology while seeking international cooperation to address risks associated with increasingly powerful AI systems.

As the United States and China intensify their competition for technological leadership, Thursday’s Trump-Xi meeting could provide an opportunity to establish whether cooperation on AI safety can advance alongside increasingly restrictive semiconductor policies.

Related Posts