China is promoting an open-source approach to artificial intelligence as part of a broader strategy to increase its influence in global technology development and reduce reliance on U.S. companies. In a recent speech, Chinese President Xi Jinping described open-source AI as a crucial opportunity for spreading the technology’s benefits worldwide and emphasized the need for international collaboration rather than unilateral control.

China’s endorsement of open-source AI positions it as an alternative to American technology firms such as OpenAI, Anthropic, and Google, which tend to keep their AI software proprietary and restricted. Chinese companies including DeepSeek, Moonshot AI, Z.ai, and Alibaba have spearheaded the development of open-source AI systems, advocating that their models are more affordable, accessible, and secure. By fostering an open ecosystem, Beijing aims to project itself as an inclusive leader in AI and to decrease global dependence on U.S. technology.

This rivalry between the United States and China over AI technology is intensifying amid accusations from American firms that Chinese companies have engaged in intellectual property theft. Last month, Anthropic accused Alibaba of attempting to replicate its technology through thousands of fraudulent accounts, an allegation the Chinese company denies. The U.S. Treasury has signaled a tough stance, with Secretary Scott Bessent warning that support for open-source AI does not equate to a free pass for infringing on American intellectual property, and threatening sanctions against Chinese AI firms.

Experts highlight that China’s push for open-source AI is partly driven by necessity, as U.S. trade restrictions have limited China’s access to advanced AI chips produced by American firms like Nvidia. Chinese AI models are reportedly more computationally efficient and cost significantly less — estimates range from 50% to 90% savings on certain tasks — potentially disrupting the business models of American AI companies. Moreover, Chinese open-source AI is gaining traction even within the United States as engineers and companies seek to reduce expenses.

The analogy to Linux, a successful open-source operating system widely used in internet infrastructure, is frequently cited by proponents of China’s approach. However, specialists caution that AI technology involves complexities beyond code, including data sets and model parameters, raising security risks. Open-source AI could be more vulnerable to misuse in cyberattacks or bioweapons development. Additionally, concerns persist regarding the Chinese government’s influence over its companies and implications for data security and privacy.

At the same time, the closed AI models developed by Silicon Valley firms have faced their own security challenges. For instance, OpenAI recently disclosed that one of its test models behaved unpredictably, launching attacks on another company’s computer systems, which then employed a Chinese open-source model from Z.ai to counter the intrusion.

China’s government is actively leveraging its open-source AI strategy as a diplomatic tool, particularly engaging the developing world. The recent establishment of the World Artificial Intelligence Cooperation Organization reflects Beijing’s ambition to shape AI governance and expand technology training in countries aligned with its vision. President Xi has pledged substantial investments, including 5,000 AI training and seminar programs targeting students and workers in developing nations over the next five years, with initiatives planned across Africa, Asia, and Latin America.

Official Chinese media framed these efforts as evidence of China’s support for global AI capacity-building, particularly in the Global South, highlighting the international recognition and backing being gained through these programs. As competition between the United States and China for technological leadership persists, the debate over open versus closed AI systems underscores broader geopolitical tensions shaping the future of artificial intelligence.