China’s AI talent returns from the United States as elite students’ U.S. PhD applications fall from 50% to 20% over the past year

Financial Times reports that in Beijing, Hangzhou and Shenzhen’s tech hubs, a subtle but pivotal shift is taking place: a wave of top AI talents is leaving the United States to return to China.

In the last 12 months, more than 30 AI researchers have left the U.S. to return to China, a surge from just a few percent before. At the same time, the share of engineering students from Tsinghua University applying to study for PhDs in the U.S. fell sharply from 50% to 20% after the pandemic.

Return wave and notable departures

The trend is increasingly systemic rather than individual, raising questions about what is driving Silicon Valley’s appeal to wane as a global tech hub.

Notable people switching back include: Wu Yonghui leaving Google DeepMind to lead language model development at ByteDance; Yao Shunyu leaving OpenAI to join Tencent; Roger Jiang leaving OpenAI to start a robotics company in Shenzhen; and Zhou Hao being recruited by Alibaba from DeepMind.

Cause: China’s shift from deployment to innovation

The report links the talent movement to a broader change: China is not only deploying technology but is becoming a true AI innovation center.

Macro-level, China is accelerating AI deployment across its economy—from autonomous taxis in Beijing to automated trading in Shanghai—faster than the United States, which remains entangled in ethics debates and regulatory frictions.

Development: supply-chain and robotics ecosystem

The advantage is reinforced by a robust supply chain spanning hardware manufacturing, electric vehicles and robotics.

According to Steve Hsu, a professor at Michigan State University, Shenzhen alone hosts at least 100 humanoid robot companies, creating a practical AI lab. He is quoted saying, “If you work on robotics, you must be in Shenzhen,” highlighting speed and hardware ecosystem advantages that the U.S. cannot easily match.

Data and hiring economics

Beyond career opportunities, micro factors also contribute to the return wave. Hiring firms report that salaries for top AI researchers in China, when adjusted for tax and cost of living, have surpassed those in Silicon Valley. The report says this can translate into a higher quality of life, including housing and services, and amenities that engineers in the Bay Area may find harder to obtain.

Impact: a two-way talent flow

The talent flows are bidirectional. While some Chinese researchers return home, others move toward Western hubs when opportunity arises. The report describes this as a rebalancing rather than a one-way exodus.

Although the United States remains home to the world’s strongest startup ecosystem, the flow of talent is increasingly two-way. The report cites Meta’s recent recruitment of an AI engineer from Alibaba as evidence that when one person leaves, another arrives.

Analysis: normalization of global competition

The trend signals a maturing Chinese tech sector. China is no longer portrayed as a passive recipient of Western technology, but as an active generator of breakthroughs.

For those returning, the report says the future of AI is not solely being written in California, but will also be shaped by data from billions of devices, factories and cities in China. Over decades, the trajectory had been largely one-sided toward the West; now, with China accelerating innovation and the U.S. tightening policies, the talent flow is balancing.

The report concludes that the winners will be nations and organizations that can attract, retain and grow talent in an increasingly competitive world.

Source: Financial Times