Nvidia CEO Jensen Huang says China leads in AI research output

The manufacturing of talent

Nvidia CEO Jensen Huang recently framed the competition for artificial intelligence as a manufacturing story. Rather than focusing solely on hardware, he argues that China excels at producing the human capital required to build large-scale models. He suggested that the ability to scale a workforce of researchers is a primary competitive advantage for the Chinese market.

During a talk at GTC Berlin in October 2026, Huang offered a blunt assessment of the current landscape. He noted that nearly half of the world's AI researchers are of Chinese origin, stating, "Nobody manufactures AI researchers like China does." This perspective treats China as a peer competitor to the United States rather than a secondary follower.

Huang identified three specific areas where he believes China shows significant strength. These include the sheer number of researchers, the production of physical chips, and the development of open AI models. By highlighting these areas, he emphasizes that the technological gap is narrowing through a massive influx of human expertise.

Data on elite researchers

Recent data from the Carnegie Endowment appears to support Huang's observations regarding the shift in talent. An analysis from September 2026 tracked authors of papers for the NeurIPS 2025 conference to identify elite talent. The study found that 57% of these researchers earned their undergraduate degrees at Chinese institutions.

This figure represents a significant increase from previous years, as the percentage was only 46% in 2022. The report specifically credits Chinese universities, such as Peking and Tsinghua, with rising to prominence in training these high-level experts. This trend suggests a steady pipeline of talent moving toward Chinese academic systems.

The data also revealed a shift in where these researchers choose to work. The study found that 41% of the tracked elite researchers are currently employed in China, while only 34% are employed in the United States. This indicates that China has overtaken the US as the primary destination for the world's top AI researchers.

The impact of US policy

The rise of Chinese AI talent coincides with a tightening of US visa policies for students in STEM fields. While these regulations are often intended to protect national interests, the Carnegie data suggests they may be pushing researchers toward other countries. This creates a tension between security goals and the desire to attract global innovation.

Huang argues that walking away from global markets and people may actually weaken the American position. He suggests that by restricting access to talent, the US may be making it easier for domestic alternatives to emerge. He specifically pointed to Huawei as a company that has used the opening created by US export controls to build a stronger position in China.

Nvidia has felt these policy shifts directly on its financial statements. The company saw its share of the AI GPU market in China drop from approximately 95% to zero after export controls were implemented. Huang uses this experience to advocate for a strategy that maintains engagement with global talent rather than isolation.

Open models and safety

Beyond the workforce numbers, Huang highlighted China's work on open AI models as a specific point of strength. Open models are significant because they can spread quickly across the globe once they are released. This capability allows for rapid adoption and iteration outside of controlled corporate environments.

Despite the intense competition for market share and talent, Huang called for continued dialogue between the US and China regarding AI safety. He believes there are shared risks in the development of these technologies that neither country can manage in total isolation. This call for cooperation exists alongside his warnings about China's growing capabilities.

The situation presents a complex picture of a bifurcated tech landscape. While the US and China compete for the best minds and the largest market shares, the underlying risks of AI development remain a common concern. Huang's stance suggests that the most effective way to manage these risks is through active communication and global talent attraction.