As the United States and China compete for leadership in artificial intelligence (AI) research, China has intensified efforts to attract foreign scientists, but these initiatives have yet to yield significant results. Over the past two years, China has introduced new visa programs for scientists and offered substantial research grants to entice international talent. Despite these incentives, foreign AI researchers are not flocking to China in large numbers, according to a recent study conducted by the Carnegie Endowment for International Peace.

The study reveals that more Chinese-origin AI researchers are currently working in the United States than the inverse, indicating a continued migration of talent toward America. For every 30 Chinese researchers employed in the U.S. in 2025, only one has made the journey from the U.S. to work in China. The research team analyzed data from last year’s NeurIPS conference, a leading event in machine learning and computational neuroscience, examining authors’ educational backgrounds alongside their current workplaces to track migration patterns.

“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,” said Damien Ma, who led the study. The findings suggest that restrictions imposed by the Trump administration on foreign researchers, including tougher visa policies and cuts to research funding, may not yet have diminished the U.S.’s ability to attract Chinese scientists—a delay attributed to many researchers having moved before these policies took effect.

Meanwhile, China is experiencing a significant rise in talent retention among domestically educated researchers. The country now hosts an increasing share of AI scholars who have graduated from its own universities and choose to remain rather than pursue opportunities abroad. This shift is underscored by the growing presence of Chinese academic institutions among top contributors to AI research; between 2022 and 2025, the number of leading NeurIPS institutions based in China doubled from six to twelve. Peking University and Tsinghua University in particular have surged ahead, surpassing renowned U.S. schools such as M.I.T., Stanford, and Carnegie Mellon. Additionally, Chinese researchers accounted for 41 percent of NeurIPS authors last year, compared to 34 percent from the United States.

China’s expanding ecosystem—with well-funded universities, major technology companies, and startups—provides ample domestic opportunities for AI experts. Kevin Xu, founder of the AI-focused hedge fund Interconnected Capital, notes that the rise of influential Chinese AI firms like DeepSeek and Moonshot is making domestic positions increasingly attractive. However, the Chinese government’s restrictions on the international travel of AI entrepreneurs and researchers, including founders of high-profile startups, may also limit the mobility of Chinese talent.

The government’s control is exemplified by the case of Manus, an AI startup once poised for acquisition by Meta, whose founders were prevented from leaving China. Since then, travel by Chinese AI professionals and their families has reportedly been further restricted.

Although China continues to struggle with recruiting foreign talent, there have been notable instances of Chinese researchers returning from abroad, such as Yao Shunyu, a former OpenAI scientist who joined Chinese tech giant Tencent as chief AI scientist last year. Despite these efforts, Silicon Valley remains more attractive to many American researchers due to broader investment opportunities and cultural factors. Language barriers and the challenge of integrating into a society with fewer immigrants further complicate relocation to China.

“For everybody else who’s outside the walled garden, culturally or digitally, it’s still very hard to break into,” Xu said, highlighting the ongoing hurdles China faces in its bid to become a global AI hub.