How Chinese talent powers US AI dominance

The US and China are the two leading countries in the development of artificial intelligence (AI) worldwide. In 2025, the US government spent US$471 billion on AI development, followed by China’s US$119 billion. Despite the lower financial allocation, China was able to develop an AI-powered chatbot, DeepSeek, at a lower cost. In addition, China has recently drawn global attention with its “humanoid robots”, meaning machines that are designed to look and act like people. 

China’s larger workforce could have offset its relatively lower spending on AI development and its technological disadvantage. Data from China’s National Bureau of Statistics indicated that more than five million people worked in China’s information industry in 2023, encompassing software development, programming languages and data transmission, compared with only 2.8 million in the US, as recorded by the US Census Bureau. 

While China has a large pool of AI talent, its relatively lower financial investment limits its ability to attract and retain top-tier researchers. By contrast, the US offers a more open innovation environment and higher compensation, making it an attractive destination for Chinese AI professionals. However, escalating geopolitical tensions and national security concerns may constrain the US from continuing to rely on Chinese talent to advance its AI industry.

Compared with other sectors in China, employment in the information industry has grown much more rapidly in recent years, underscoring its dynamism and the new opportunities it has created. This expansion has likely been driven not only by China’s industrial policy shift towards AI, but also by state-led digitalisation across the economy, including the widespread adoption of digital payments and facial recognition technologies.

The abundant supply of graduates in information technology fulfilled the growing demand in the job market in China’s AI industry. China’s Ministry of Education in 2022 estimated that nearly 1.6 million college graduates studied science and engineering, accounting for 40% of the total number of college graduates that year.

Unlike China’s relatively larger share of students in science and engineering, less than 8% of US students majored in engineering technologies, as well as computer and information sciences and support services combined. Instead, the majority of students in the US (38%) majored in liberal arts and sciences, general studies, and humanities in 2021/2022, according to the National Center for Education Statistics.    

Data from the Center for Security and Emerging Technology (CSET) also showed that China had more than four times the number of graduates in Science, Technology, Engineering and Mathematics (STEM) than the US in 2020. The large number of STEM students in China has also been driven by the country’s rapid expansion of AI-related programmes in higher education, with 535 universities now offering specialised AI majors, and the government relying on their growing STEM talent pool to support its AI research and development ambition.

Foreign students have filled the STEM fields in the US. In the 2024/2025 academic year, the total number of international students at US colleges and universities reached an all-time high of 1.17 million, or 6% of the total US higher education population. Around 57% of international students in the US were majoring in STEM in the same year, according to the Institute of International Education. 

Many Chinese students pursued their advanced studies in the US. In the 2024/2025 academic year, Chinese students accounted for 23% of total foreign students in the US, behind India’s 31%. Chinese students are especially interested in studying engineering and mathematics/computer science in the US, accounting for 22% and 21% of international students in these fields respectively, behind India’s shares of 38% and 52%.

Many foreign students stay in the US for work after graduation, contributing to the development of AI technology there. Between 2018 and 2022, 77% of non-American AI researchers remained in the US after completing their PhD studies in the country. According to the Global AI talent tracker, 80% of AI talent educated in the US stayed for work after graduating, in sharp contrast to only 11% in China. The same data source also shows that Chinese talent is the largest source of top-tier AI researchers in the US. In 2022, 38% of top researchers in US institutions were originally from China.

Data from the US Department of Labor shows a similar reliance on foreign workers in AI-related industries. Foreign-born workers accounted for 27% of total workers in computer and mathematical occupations in 2024, up from 20% twenty years ago.

Compared with other professional occupations in the US, there is a relatively higher percentage of foreign-born workers in computer and mathematical fields. This also reflects the fact that most American students do not major in STEM fields. Hence, computer and mathematical occupations are largely filled by foreign nationals. The US government could encourage more local students to study STEM, but this will take time. Therefore, US demand for AI talent from overseas is likely to continue growing alongside the industry’s rapid development.

China’s greater number of people studying and working in the Chinese AI industry could provide the country with an advantage in catching up with America’s level of technology. Yet, the excessive government intervention in AI development could restrict private businesses’ freedom to innovate.

The Chinese government could offer better remuneration and benefits to encourage Chinese and international talent to work in China. However, history shows that financial incentives do not always guarantee success. For example, China has not developed a globally competitive semiconductor industry despite efforts to poach foreign talent, particularly from Taiwan, over the past two decades.

Despite its more liberal environment, which attracts top AI talent worldwide, the Trump administration’s stricter policy towards foreign workers is a disadvantage for the country’s future AI development. In particular, to prevent Chinese students and scholars from transferring sensitive technologies to China, the US has been more cautious in issuing visas to Chinese nationals. The leakage of business secrets or technology-related information by a single individual could potentially cause significant damage to the US AI industry.

AI talent from other countries and regions, such as India, Singapore, Taiwan and South Korea, could compensate for the potential shortfall of Chinese AI talent. Due to the insufficient number of domestic students and workforce in engineering and computer sciences to meet the market demand, hiring foreign talent remains essential in keeping America’s AI industry competitive. The reliance on foreign talent means that the US will have to enhance its relations with its allies in constructing a resilient AI ecosystem.

Related: Who steers AI: China’s industrial state vs America’s frontier builders? | When cost and practical application takes priority: China surpasses US in AI adoption