Illustration: Chen Xia/GT
The increasing use of Chinese open-source artificial intelligence (AI) models by US technology and AI companies has received significant attention and sparked heated debate within the US industry.
As NBC News reported on Sunday, growing support for China’s AI models could spell trouble for the US AI industry, a phenomenon that has raised questions about whether the US’s pursuit of closed models is misguided, with some experts even calling the US’ lack of a strong open source model an “existential” threat to democracy.
Western media is increasingly paying attention to Silicon Valley’s focus on Chinese-made AI tools. A Bloomberg article last month was headlined, “How Much of Silicon Valley Is Built with Chinese AI?” He pointed out that the Chinese model has overtaken the US model in cumulative downloads by developers, that Alibaba Qwen has surpassed Metaplatform’s Llama in downloads, and that derivative systems built on Qwen account for more than 40% of new language models.
Another article from Bloomberg considers whether “deep-seek moments” are becoming the new normal for AI.
These changes are no coincidence. It stems from market rationality that intersects with the different development paths of the US and Chinese AI industries. Fundamentally, business decisions prioritize technical performance, cost efficiency, and development agility. China’s open source model is an attractive solution, especially for coding and software development entrepreneurs who need to ship products quickly. They compete with top international models in functionality while offering a decisive cost advantage.
Additionally, its open-source nature, which allows for download, tweaking, and local deployment, inherently addresses data privacy and content security concerns.
Analytical data further illuminates this dynamic. The Chinese tech giant’s total capital spending from 2023 to 2025 was 82% lower than its U.S. peers, but the performance gap between its leading models has become much smaller, according to Jefferies analysts cited by Bloomberg. This marriage of low cost and high performance accurately describes the pragmatic choices made in Silicon Valley and highlights the growing global appeal of China’s open source approach.
A notable example comes from Brian Chesky, co-founder and CEO of San Francisco-based online accommodation booking giant Airbnb, who openly stated that the company is “heavily relying” on Alibaba’s Qwen model to power its AI-driven customer service agents. This actual adoption clearly shows that developers are “voting” on the code, which is more convincing than any performance ranking. This proves that market valuation ultimately depends on the practical value and affordability of the technology.
This market-driven reality reflects the increasingly clear trajectory of AI development in China and the United States. Leading US companies such as OpenAI are focused on progress sustained by closed-source, high-margin business models. In contrast, major Chinese AI companies such as DeepSeek have adopted an open source philosophy centered on widespread adoption. While pursuing technological excellence, they are committed to lowering the barrier to application through open collaboration and accelerating AI applications into manufacturing, services, and even daily life faster.
Against this backdrop, some voices in the United States are concerned about this development, seeing it as an erosion of America’s technological superiority. Part of this concern stems from anxiety about the potential for change in technology leadership, which is further amplified by zero-sum thinking.
However, it is important to recognize that over-politicizing technology flows can pose a major threat to your country’s innovation ecosystem. Past strategies aimed at containing Chinese innovation, from chip restrictions to various technological barriers, have frequently stimulated and accelerated China’s independence, while also limiting opportunities for American companies. The AI sector does not need to repeat this counterproductive pattern.
China’s progress in open source AI is the result of persistent open innovation despite external pressure. This proves to the world that the true value of technological advancement lies in connection and collaboration. For the international community, the greatest risk lies not in the origins of the technology, but in closing the door to cooperation, fragmenting global innovation networks, and slowing the historic process by which AI will empower all humanity.
