Video: Timo Lenzen
“I S IT POSSIBLE that the United States falls behind China?” Jensen Huang, the boss of Nvidia, asked himself during a question-and-answer session about artificial intelligence late last year. “The answer is absolutely yes.” That may seem surprising—for much of the past decade America has been comfortably ahead in the AI race, home to the most advanced companies producing frontier models. Its engineers have access to deep pools of capital as well as a regular supply of Nvidia’s cutting-edge chips. But Mr Huang’s concern related to an equally important ingredient of innovation: human talent.
Countries ranked by number of top AI researchers*
United States
China
Switzerland Spain South Korea Singapore Russia Portugal Norway Netherlands NA Japan Italy Israel India Greece Germany France Finland Denmark Canada Britain Belgium Austria Australia
United States
China
1st 1st 5th 5th 10th 10th 15th 15th 20th 20th 25th 25th 30th 30th 35th 35th 40th 40th 45th 45th 2016 2019 2022 2025 Total Until recently, most leading AI research was produced by experts based in the West. That is changing. In 2025, for the first time, more studies presented at the world’s top AI conference had lead authors based in China than in either America or Europe. To better understand the international ebbs and flows of AI talent, The Economist tracked the education histories of researchers who presented papers at the December 2025 edition of the Conference on Neural Information Processing Systems (Neur IPS), the world’s largest and most prestigious AI gathering. More than 21,000 papers were submitted to Neur IPS, of which roughly a quarter were accepted. MacroPolo, a now-shuttered think tank, analysed the change in educational background of Neur IPS authors in 2019 and 2022. We applied their method to a random sample of 600 papers (authored by almost 4,000 researchers) from 2025. placeholder In 2019, 29% of AI researchers who presented at Neur IPS began their careers in China.By 2025 half did. Over the same period, the share who started out in America fell from 20% to 12%.Moreover, fewer Chinese undergraduates leave upon graduation. By 2025 three fifths of AI researchers continued their studies in China.placeholder Nine of the top ten institutions where authors from the 2025 conference earned their undergraduate degrees were in China. Graduates of Tsinghua University alone accounted for 4% of researchers at Neur IPS. MIT, the leading American institution, produced 1%. The analysis also shows the extent to which America’s AI efforts rely on Chinese-born researchers. Among authors affiliated with American institutions, roughly 35% have a Chinese undergraduate degree (as many as have an American one). That being said, Neur IPS may not be entirely representative of the field. Chinese researchers might feel stronger incentives to present at the conference: to win promotions at academic institutions, for example, scientists often need top conference papers on their CV. What’s more, China’s culture of open-source models may encourage its authors to publish in academic forums, whereas America’s leading talent is increasingly concentrated in secretive frontier labs. There are other measures by which the importance of Chinese researchers to America can be gauged. When Meta, a tech company, announced the researchers staffing its new “superintelligence lab”, in June, a leaked list revealed that half were described as being from China. The Economist ’s analysis of 483 contributors to Open AI ’s GPT -5 (which includes AI researchers as well as marketing, design and leadership staff) found that 15% had at least one degree from a Chinese institution.
Student
Early career
Mid-career
Late-career
China is increasingly holding on to its AI talent. According to Digital Science, a data firm, China now has more active AI researchers than America, Britain and Europe combined—though it still trails the West per head of population. What’s more, China’s cohort skews younger: 47% are students, compared with about 30% in the West. The country also prioritises education in science, technology, engineering and maths (STEM): around two-fifths of Chinese university students study STEM subjects, roughly double America’s share. Not all of these graduates will produce frontier innovations, but scale matters. A large pool of AI -savvy researchers increases the chance of breakthroughs and means new technologies spread faster. “China is creating this high-quality, highly trained workforce who is AI -sensitive,” says Daniel Hook, the boss of Digital Science. “That’s just going to mean so many companies coming out of China in the future.” Chinese boffins are increasingly choosing to stay in the country. In 2019 roughly a third of Neur IPS authors who had completed their undergraduate degrees in China remained there. By 2022 that share had risen to 58%; in 2025 it reached 68%. Some of the country’s best innovations have come from entirely home-grown talent—none of the core contributors to DeepSeek R 1, a Chinese model that stunned rivals when it was released in January 2025, held degrees from outside of China. These changes reflect both pull and push. Chinese universities are increasingly ranked among the best in the world. At the same time, initiatives to lure talented researchers back to China, such as the Qiming Plan, offer salaries of more than 700,000 yuan ($100,000), generous research grants and help with housing. At the same time America has become a less attractive destination. Funding cuts and visa uncertainty have unsettled would-be applicants, as has increasing suspicion of their loyalties. Last year Purdue University rescinded offers to more than 100 graduate students, most of them Chinese, after being asked by lawmakers to document researchers’ ties to institutions in China. At American AI meetings some Chinese researchers feel the need to clarify they are not corporate spies.
Now working:
Inside China
Outside China
More are therefore heading home. In 2019 just 12% of Chinese Neur IPS researchers who had earned graduate degrees abroad had returned to China. By 2025 that share had more than doubled to 28%. The Economist spoke with Chinese-born early-career researchers who have recently relocated back home from America, or have moved back and forth between the two countries. Some still consider America to have a stronger research environment or complain of fierce competition and long hours at China’s fast-growing firms. Yet they said on balance a strong job market, interesting opportunities and proximity to family now outweigh those drawbacks. America’s appeal has not vanished. It still draws more international talent than anywhere else and most Chinese researchers who complete graduate degrees in America stay on to work. Following up on a sample of Chinese-born, America-based Neur IPS authors from the 2019 conference, 87% were still there in 2025. “Long-standing institutions just don’t disappear overnight,” says Matt Sheehan, of the Carnegie Endowment for International Peace, who performed the research and worked on the original MacroPolo analysis. But the numbers increasingly favour China. Using the authors of Neur IPS papers as a metric, around 37% of the world’s top AI researchers now work in Chinese organisations, compared with 32% in American ones. If the trend of the past decade continues, by 2028 top Chinese-based researchers could outnumber American-based ones by two to one. According to Mr Huang, for a country to lead in AI “winning developers is everything”. The battle for talent looks increasingly one-sided.■
논증 분석
유형: diagnostic
핵심 주장
중국은 AI 인재 경쟁에서 미국을 앞서가고 있으며, 이는 AI 혁신의 핵심 요소인 인적 자원 측면에서 미국의 AI 패권에 심각한 위협이 되고 있다.
논리구조
- 전제: Nvidia의 CEO Jensen Huang이 스스로 인정했듯이, 미국이 China에 뒤처질 가능성은 ‘절대적으로 그렇다’—AI 혁신에서 자본과 칩 못지않게 인적 자원이 결정적 요소다.
- 진단: 최상위 AI 연구자 수를 기준으로 국가별 순위를 분석한 결과, China가 보유한 최고 수준의 AI 연구 인재 수에서 United States를 추월하고 있다.
- 논거: 지난 10년간 United States는 최첨단 프론티어 모델을 개발하는 선도 기업들을 보유하며 AI 경쟁에서 편안하게 앞서 있었으나, 인재 풀의 균형이 변화하고 있다.
- 진단: AI 연구자 인재 경쟁에서 China의 부상은 단순히 반도체 접근성이나 자본 규모의 문제가 아니라, 근본적인 인적 혁신 역량의 이동을 의미한다.
결론
China의 AI 인재 우위는 United States의 AI 기술 패권에 구조적 위협이 되며, Jensen Huang의 경고처럼 미국이 AI 경쟁에서 뒤처질 현실적 가능성이 인재 차원에서 이미 현실화되고 있다.