China is keeping a larger share of its own AI researchers, but that is not the same as attracting more researchers from abroad. A 2026 Carnegie Endowment analysis of NeurIPS authors finds that the U.S.–China talent balance still favors the United States, even as China’s research ecosystem becomes a more credible place for Chinese-origin researchers to work.
What Carnegie’s numbers show
Carnegie compared the education histories and current workplaces of authors whose papers were accepted at NeurIPS, a leading AI research conference. Its 2025 dataset includes 5,823 accepted papers and 25,677 unique authors. Career-flow figures draw on the 10,280 authors—40 percent of that cohort—for whom Carnegie had complete information on undergraduate origin, graduate education, and current workplace.
As an Amazon Associate I earn from qualifying purchases.
The figures describe a conference-based group, not all AI researchers. Carnegie says broader sample checks preserve its main U.S.–China findings, although some results for other regions are sensitive to the smaller complete-history sample.
Origin and workplace are different measures
In this analysis, “Chinese-origin” means a researcher earned an undergraduate degree in China. It does not necessarily mean the researcher’s nationality or birthplace. Carnegie classifies workplace by the recorded institution or employer; for a multinational company, that may be the company’s home country rather than the researcher’s physical office.
#1 Best Overall
| Measure | Carnegie’s 2025 finding | What it means |
|---|---|---|
| Undergraduate origin | 57% of the NeurIPS author cohort had undergraduate degrees from China, 11 percentage points higher than in 2022. | China is a major source of researchers in this cohort. |
| Current workplace | 41% of the sampled AI talent worked in China; 34% worked in the United States. | These are shares of the sampled talent by workplace, not shares of all AI workers in either country. |
| Chinese-origin researchers working in China | 69% in 2025, compared with 57% in 2022. | A larger share of researchers with Chinese undergraduate degrees were working in China. |
| Chinese-origin researchers working in the United States | Their share of the researchers working in the United States rose by 4 percentage points from 2022 to 2025. | The U.S. still hosted Chinese-origin researchers, and their share of its sampled AI workforce increased. |
Does the 30-to-1 figure mean researchers moved in 2025?
No. Carnegie describes a 30-to-1 balance in talent flows between the United States and China, but calculates that comparison from researchers’ undergraduate origin and current workplace. It is not a count of people who moved between the countries during 2025, nor does it establish when any move happened. The figure captures a long-run pattern in where people studied and where they work now.
The distinction matters: a Chinese-origin researcher currently employed in the United States contributes to the origin-to-workplace comparison, whether that person moved last year or many years ago. Carnegie’s results show the U.S.–China flow remains strongly tilted toward the United States, not that China saw a measured wave of departures in 2025.
Why China can retain researchers without recruiting many from abroad
More opportunities to do consequential work at home
Carnegie’s reporting describes a stronger domestic AI ecosystem in China, spanning universities, major technology companies, and startups. DeepSeek and Moonshot are cited as examples of companies that have raised the profile of local research opportunities. Tencent’s appointment of former OpenAI researcher Yao Shunyu as chief AI scientist is another example of a prominent role being created within China.
Recommended Free Tools
Those examples help explain why Chinese-origin researchers may find compelling work at home. They do not measure the number of foreign researchers hired, or establish why any particular scientist chose one country over another.
Attracting people from abroad is a separate challenge
Carnegie’s reporting points to the continuing appeal of U.S. academic and startup opportunities and research funding. It also notes that relocating to a country with a different language and relatively few immigrants can be difficult. The analysis raises tighter U.S. visa conditions, domestic industry opportunities, and changing geopolitical conditions as factors that could shape whether Chinese-origin researchers move to the United States or stay in China.
These are contextual explanations, not a causal ranking. The cited study does not establish how much compensation, visa policy, language, research funding, or geopolitics affects researchers’ individual decisions. Its reporting mentions a scientist visa and research grants but does not provide the visa’s formal name, eligibility requirements, start date, or grant amounts.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What the trend says—and what it does not
China’s gains in this cohort are clearest in the distinction between developing talent and attracting it from elsewhere. A larger share of NeurIPS authors had undergraduate degrees from China than in 2022, and a larger share of Chinese-origin researchers were working in China. At the same time, Carnegie’s origin-to-workplace comparison remains substantially in the United States’ favor.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsDamien Ma, who led the study, said: “Despite U.S.-China tensions, the number of Chinese-origin researchers working in the U.S. actually rose by four percentage points rather than declined” compared with three years earlier. That finding sits alongside the increase in the share of Chinese-origin researchers working in China: the two trends are not mutually exclusive.
The results concern NeurIPS authors with recorded career histories. They indicate patterns within a selective conference cohort, not a complete census of global AI talent or proof that a specific policy or workplace factor caused the changes.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
