China leads global AI development with its open-weight models

Renowned Chinese artificial intelligence scholar Wang Jian said it is too simplistic to frame AI development as a matter of China-United States competition, arguing that the more important question is which country can contribute more to the world through AI.

Wang, a member of the National Committee of the Chinese People's Political Consultative Conference and an academician of the Chinese Academy of Engineering, made the remarks in a recent interview on the sidelines of the ongoing annual two sessions of the top legislature and political advisory body.

"Many people may be concerned about the race between China and the US in AI development, but I have always felt that, strictly speaking ... what really matters is which country can make a greater contribution to the world through AI," said Wang, director of Zhejiang Lab and founder of Alibaba Group Holding's cloud computing unit.

"In particular, because we have an open-source approach, I have always believed that China is in a position to make the greatest contribution to global AI development," he said.

Wang said that what is often called "open source" in this context is more accurately described as open weights — making model weights available to others — and that this reflects a level of resource commitment far beyond what many imagine.

"When you develop a large language model and open up its weights, what you are really opening to others is the computing power, and even the electricity, that has been consumed behind it," he said. "Its significance goes far beyond simply making code available."

"When code is opened, what lies behind it is mainly a programmer's intellectual effort. That creativity is priceless, of course, but the actual resources consumed may not be as large as people imagine," Wang said.

"With large models, however, opening them up may mean that huge sums of money have already been spent, massive amounts of electricity have been used, and enormous computing resources have already been consumed," he added.

"That is why I believe that when China makes its large AI models open and turns them into open-weight models, the significance goes beyond the traditional logic of open source," the expert said. "In a sense, we are defining how this new stage of development should proceed."

Wang noted that as China's science and technology advances to a world-leading level, it will not only benefit the country itself but also help countries in the Global South and other developing nations.

The draft Government Work Report submitted to the ongoing annual session of the National People's Congress stressed "support for the development of opensource AI communities and building a vibrant open-source ecosystem".

A joint study by the Massachusetts Institute of Technology and open-source AI startup Hugging Face last year found that Chinese open-source models achieved a download share of 17.1 percent between August 2024 and August 2025, surpassing the US for the first time, which had 15.8 percent. Most of the downloads for Chinese models came from DeepSeek and Alibaba's Qwen.

According to The Associated Press, a report released by Microsoft in January found that the rise of Chinese startup DeepSeek has fueled wider AI adoption across the developing world, given its free and open source models.

In many African countries, including Ethiopia, Zimbabwe, Uganda and Niger, DeepSeek's market share was between 11 percent and 14 percent, the report said.

During the interview, Wang also highlighted that China's approach to developing AI focuses on the technology's broader role in driving social development, which he believes can in turn advance the technology itself.

The draft Government Work Report stresses the need to "take more effective measures to facilitate employment and entrepreneurship through better adaptation to the development of AI technologies".

"This is about how AI can better promote employment and expand its contribution to the broader economy," Wang said.

From this perspective, China's thinking has gone far beyond the AI technologies themselves and their specific application scenarios, placing greater emphasis on AI's broader role in driving social development, he said.

According to Wang, once AI is able to boost employment and stimulate growth in more sectors, it will in turn help promote technological progress.

He cited the invention of paper and electricity as examples, saying both gave rise to many new products and types of jobs.

"Technology does not directly create companies or professions, but creative people can use new technologies to create many new resources," he said.