China’s AI industry is entering a new phase, and the appointment of former OpenAI researcher Yao Shunyu as chief AI scientist at Tencent highlights the shift.
For years, the most ambitious talk about artificial general intelligence, or AGI, came primarily from U.S. labs such as OpenAI, Anthropic, and Google DeepMind. Chinese companies, constrained by U.S. chip restrictions and focused on commercialization, generally concentrated on practical AI applications in manufacturing, consumer devices, and internet services rather than near-term AGI ambitions.
That distinction is beginning to blur.
Register for Tekedia Mini-MBA edition 20 (June 8 – Sept 5, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab.
Yao’s goal: a long-term AGI organization in China
Speaking at a Tencent event in Beijing, Yao said he wants to build a long-term organization dedicated to AGI research in China.
“My personal goal is that in China we should establish a long-term AGI organization,” Yao said.
Yao argued that achieving AGI will require advances in foundational research, product development, and frontier exploration, not just incremental improvements to existing applications.
He also suggested that the future AI market will be far larger than any single product category.
“I don’t think ChatGPT or Claude will be the only super-app,” Yao said, adding that the opportunity is in the “trillions of dollars.”
His emphasis on performance and cost reflects a distinctly Chinese approach to AI development: smaller, more efficient models that can be deployed at scale across real-world applications.
A notable contrast with U.S. caution
Yao’s optimism comes as some U.S. AI leaders are sounding more cautious about the pace of frontier AI development.
On Thursday, Anthropic warned that advanced models are approaching the point where they could improve themselves without direct human oversight. The company called for a slowdown or pause in new model development to reduce the risk of societal disruption.
Anthropic has consistently positioned AI safety as a central concern, though critics argue that some of its warnings also have competitive implications in the race against rival AI labs.
Why this matters
Yao’s move from OpenAI to Tencent illustrates a broader trend: China is increasingly attracting back researchers who trained at leading U.S. AI organizations.
That talent flow is becoming a strategic issue in the U.S.-China technology competition.
Several factors are contributing to the shift:
- China is investing heavily in AI and basic research.
- Beijing has made scientific and technological self-reliance a national priority and is increasing funding for foundational research over the next five years.
- U.S. immigration uncertainty
- Some Chinese researchers have become less certain about long-term career prospects in the United States and are choosing to return home, even if compensation may be lower.
- Chinese firms are becoming more ambitious.
- Major companies are no longer content to focus only on applications. They increasingly want to compete in frontier model research.
Trend
Recent talent moves
- Alibaba reportedly hired former Google DeepMind researcher Hao Zhou to support the development of its Qwen AI models.
- Former Google DeepMind executive Wu Yonghui left to lead research at ByteDance Seed in 2025.
- That move reflected the growing pull of Chinese AI labs for senior research talent.
- Startup Moonshot AI was founded by Yang Zhilin, who previously worked at Meta AI and Google Brain.
- The company is now one of China’s more closely watched AI startups.
The bigger picture
The key takeaway is not that China has suddenly overtaken the United States in AGI research. U.S. companies still lead many frontier model benchmarks and benefit from stronger access to advanced chips and capital.
What is changing is the strategic ambition of Chinese AI firms.
Instead of treating AGI as a distant, mostly Western research project, companies such as Tencent are beginning to frame it as a long-term national and corporate objective. The arrival of researchers with experience at OpenAI and other frontier labs gives those efforts more credibility and technical depth.
However, it not clear whether that results in a genuine Chinese AGI organization. But the talent migration and the change in rhetoric suggest that the U.S.-China AI competition is evolving from a contest over applications and deployment into a contest over the most advanced forms of AI itself.