The last few months have seen OpenAI hemorrhage resources mainly to Meta, which has actively poached some of its top people, but it’s now losing researchers to Chinese labs as well.
China’s Tencent has recruited a prominent AI researcher from OpenAI in what is one of the high-profile defections of AI talent from the US to China. Yao Shunyu was a Research Scientist at OpenAI, who had worked on DeepResearch and techniques to build universal AI agent to interact with the digital world. Yao Shunyu has reportedly been hired at a pay package of 100 million yuan ($14 million), but Tencent has dismissed this news as a rumour.
Shunyu will now reportedly work on integrating AI into Tencent’s services. Tencent runs the popular WeChat platform in China, which apart from being a messenger serves as an app for payments, shopping, ride hailing and other services.
Yao Shunyu is Chinese, and had graduated from Tsinghua University in 2019 with a degree in Computer Science. He’d then pursued a PhD at Princeton. He’s previously interned at Google and OpenAI, and had been with OpenAI for nearly two years.
His move to China is yet another sign of China’s growing influence in the world of AI. Thus far, US companies employed large numbers of Chinese and Chinese-origin researchers who came to the US in search of better salaries and better opportunities. But the AI scene in China has been exploding in recent quarters, with companies like DeepSeek, Alibaba, Bytedance, and startups like Moonshot AI and Z.ai all releasing extremely capable models. In some areas, these models are better than their American counterparts — many Chinese open-source models are better than GPT-OSS, the best open-source model released by the US, and Bytedance’s Seedream 4.0 had recently gone past Google’s Nano Banana to become the top image model on Artificial Analysis’s leaderboards. As such, there are plenty of opportunities to do cutting-edge AI work in China, and it seems that Chinese researchers are beginning to pick them over US labs.
This could help accelerate the growth of China’s AI capabilities. These researchers have finished their higher education in the US, and have worked at US companies, and will take their learnings back to China. This expertise, along with China’s own progress, would help China catch up even faster in the AI race. China has other structural advantages in AI too, with much more electricity production, and a more benign regulatory environment. It remains to be seen if China can make the most of these advantages, but it seems that the stage is being set for it to eventually go past the US in AI capabilities.