Two charts are actively being discussed on Chinese social media, sparking debate over whether China has surpassed the United States in artificial intelligence. The first chart shows that by 2025, China has become the leading employer of elite AI researchers, according to data on the institutional affiliation of the first author in NeurIPS. The second chart indicates that 47% of the world's top AI researchers received their bachelor's degree from Chinese universities.
However, these charts do not provide a complete picture. NeurIPS authorship data reflects only research output from Chinese institutions but does not cover the full spectrum of AI capabilities, including commercial implementation, large-scale systems engineering, fundamental innovations, and the ability to attract global talent. Similarly, MacroPolo data, which points to the Chinese origin of 47% of researchers, suggests that many of them work outside of China, primarily in the United States, after obtaining advanced degrees abroad and joining companies such as OpenAI, Google DeepMind, Anthropic, and Meta.
A fundamental structural difference between the two AI ecosystems has become evident. China demonstrates superiority in mass talent production, possessing a vast engineering workforce and the ability to rapidly convert engineering developments into products in manufacturing, autonomous driving, robotics, and AI applications. The US advantage lies in attracting global talent, the depth of fundamental research, the development of cutting-edge models, advanced chip design, venture capital density, and the ability to concentrate the world's best minds into a single innovative system, regardless of nationality.
More significant than any single ranking are five questions. Firstly, does winning in one metric equate to winning the system overall? Although China leads in the volume of AI researcher graduates, this advantage depends on continuous investment in higher education and research infrastructure. Secondly, how is talent migration occurring? There is a trend where Chinese doctoral students educated in the US begin returning to domestic AI companies, gradually shifting the balance toward a more bilateral flow. Thirdly, is the competition defined by model benchmarks or ecosystem breadth? China leads in industrial AI integration and application deployment, while the US leads in breakthrough developments of advanced models and AI infrastructure. Fourthly, what role do access to computing power and energy play? Export controls on advanced chips create a structural limitation for AI development in China that cannot be overcome solely through talent. And fifthly, can either system independently sustain the full cycle—from education and research to development, implementation, and commercialization?
Instead of declaring a winner, it is more productive to view the situation as follows: the US and China have formed complementary strengths in different parts of the AI value chain. The ultimate determining factor in the competition is not which country has more researchers this year, but which one can build a self-sustaining system that ensures talent production, attracts global personnel, generates fundamental innovations, converts research into products, and creates economic value on a large scale. On this measure, both systems possess unique strengths and vulnerabilities.
China is traditionally considered the world leader in the mining and processing of rare earth elements (REEs). Globally, about 70% of the mining and approximately 90% of the processing of these elements takes place in China. However, a new study has revealed another side of this coin.
According to the study, while China leads in raw material extraction and processing, it lags behind the US and Japan in the most critical and expensive technologies related to rare earth elements. Data from the South China Morning Post in Hong Kong indicates that many key patents concerning advanced functional materials made from rare earth elements remain owned by Japan and America.
Rare earth elements are critically important not only for business but also for strategic and defense sectors, as they are used in many high-tech devices, including radars, sonar systems, missiles, fighters, as well as smartphones, satellites, and electric vehicles.
The research showed that over 80% of patents related to permanent magnets, catalysts, luminescent, and polishing materials made from processed rare earth elements are registered worldwide. In this context, Japan holds a leading position in permanent magnet technology, while the US dominates in catalyst, luminescent, and polishing material technologies.
It was found that China has achieved advantages only in limited areas. It still lags behind Japan and the US in several crucial manufacturing processes and material systems. This means that China has not yet been able to fully transform its raw material advantage into high-value technologies. Despite China filing a large number of patents, the international influence of its high value-added patents is less than that of its competitors. Experts emphasize that to achieve complete global leadership in the rare earth element sector, China must not only dominate mining and processing but also catch up with the US and Japan in advanced innovation and technology.
