US AI leadership built on Chinese talent faces growing pressure from visa restrictions, according to a Carnegie China Center report. Its findings connect American research gains with an increasingly important Chinese talent pool.
The United States gained a net 2,145 AI researchers in 2025, while China recorded a net loss of 1,729. The report warns that tighter visas could weaken the talent flows supporting American competitiveness.
AI Leadership Built on Cross-Border Research
The report’s talent findings identify China as the largest source of elite AI talent. Chinese-origin researchers remain central to the competition between the two countries.
Among the world’s top undergraduate AI talent, the share of Chinese origin increased from 46% in 2022 to 57% in 2025. The proportion of US-born talent fell to 13%, according to the report.
The destination figures also changed. Among Chinese-origin AI researchers, the share choosing to work in China rose from 57% in 2022 to 69% in 2025.
The report links that movement with the Trump administration’s visa tightening. It argues that restrictions are eroding a foundation of American AI competitiveness, even as the country seeks to maintain its lead.
Those findings sit alongside broader questions about how organizations plan for technology. Related coverage examines strategic thinking in leadership as operating demands evolve.
Visa Policy Conflicts With Talent Needs
Trump has repeatedly said that “we are leading China in AI,” according to the source. The report argues that restricting Chinese talent runs counter to that objective.
Its analysis describes a longstanding talent pipeline from China to the United States. Without that supply, it says America’s lead “would not be as pronounced” and questions whether it would exist.
The report frames AI competition primarily as a contest for talent. It argues that AI leadership built on global talent can continue if top researchers keep arriving, regardless of their origin.
That argument depends on maintaining an open talent-admission policy. The report says the tension between immigration restrictions and research needs will deepen if the United States continues limiting inflows.
Businesses have also opposed restrictions involving Chinese AI technology. In July 2026, nearly 200 US startups signed a letter opposing threatened sanctions and limits on using Chinese open-source models.
The source presents that opposition as evidence of concern within America’s innovation ecosystem. Separate reporting on venture-backed startup risks considers another challenge facing young businesses.
Collaboration Remains Central to the Report
The source points to earlier scientific exchanges as examples of cross-border cooperation. In the 1970s, Herbert Simon and other American AI pioneers traveled to China to lecture.
It also describes collaboration between researchers and corporate teams during the early 2000s deep-learning boom. Those teams explored technical approaches across the Pacific.
The report argues that chip export controls, compute restrictions and investment bans reduce room for scientific cooperation. It warns that such barriers may weaken America’s position in the talent contest.
For readers considering how institutions develop future leaders, related coverage describes a finance and leadership academy for teenagers.
The Carnegie China Center’s stated implication is specific: American talent policy must align with its AI ambitions. Restricting Chinese researchers, the report argues, undermines a critical source of that competitiveness.
