Carnegie China: China passed the US as the top destination for elite AI talent in 2025, rising from 2022's 27.1% share to 40.6%; the US fell from 46.4% to 34.2%
Context & Ripple Effects
A 2024 study found that China produced almost half of the world’s top AI researchers, making the country’s research-training base materially larger than the U.S. share in that analysis. A 2025 warning that U.S. AI-talent inflows had reached parity with outflows framed destination competitiveness as a developing constraint rather than a settled loss of attraction.
Carnegie China’s figures turn that underlying supply and mobility story into a location shift: the largest share of its elite-researcher sample is working in China rather than the U.S. Public discussion of the tracker also distinguishes destination share from net migration, arguing that the U.S. still posts a net gain in the sampled researchers.
First-order effects
- China-based AI universities, labs and companies gain access to the largest work-location share of Carnegie China’s elite-talent sample, while the U.S. share contracts.
- The result strengthens the link between China’s large domestic pipeline of top AI researchers and its own research employers rather than treating training and employment as separate advantages.
Second-order effects
- U.S. universities and AI employers face greater pressure to retain and recruit researchers as the destination lead flagged in earlier talent-flow analysis gives way to a smaller share of the Carnegie sample.
- China-based employers can draw more of their advanced-research hiring from researchers trained and working domestically, reducing the extent to which overseas placement is the primary route into elite AI work.
Third-order effects
- If the shift persists, frontier AI research capacity will be distributed less around a single U.S.-centered destination and more around competing national research ecosystems.
- Talent location becomes a more important component of AI sovereignty: training capacity, research institutions and employer opportunity increasingly reinforce one another within the same country.
The trend: AI competition is broadening from a contest to attract individual researchers into a contest to build self-reinforcing national talent ecosystems.