U.S. AI Industry Still Relies on Chinese Models and Manufacturing Despite Strategic Rivalry

BEIJING — The United States and China remain deeply connected through the artificial intelligence industry, even as Washington tightens scrutiny of Chinese technology and both countries compete for leadership in advanced computing.

Chinese AI models and manufacturing capabilities continue to serve some American businesses, illustrating the difficulty of separating the two economies in a technology sector that depends on software, hardware, suppliers and large-scale infrastructure.

In a report published on October 11, The Washington Post examined how Chinese AI company Moonshot and its Kimi model have found users among American businesses seeking lower-cost AI tools. The report also described the role of Chinese suppliers in producing equipment and consumer devices linked to the wider AI ecosystem.

Chinese companies such as Moonshot, DeepSeek and Alibaba have developed open-weight models that users can customize and deploy for particular tasks. These models can be attractive to businesses because they offer flexibility and may cost less to use than some proprietary alternatives.

The Washington Post reported that companies including DoorDash and Airbnb have used Chinese models for routine tasks. Such applications do not necessarily mean Chinese models outperform leading U.S. systems in every category. Instead, they illustrate how businesses evaluate AI tools according to cost, reliability and the specific job they need to perform.

The relationship extends beyond software. Chinese factories supply electronics, sensors, power equipment, cooling systems and other components used in AI-related products and data centers. The newspaper cited research estimating that U.S. hyperscalers imported roughly $45 billion worth of AI-related goods from China in 2025, although the value and share of those imports have declined as supply chains have diversified.

This dependence creates a policy dilemma for Washington. Restricting Chinese technology may reduce some security risks, but broad restrictions could also raise costs or limit access to products that American companies use in commercial applications.

U.S. officials have raised concerns about the security implications of Chinese AI models and alleged unauthorized use of American AI systems for training. Chinese companies face their own constraints, including limits on access to advanced computing chips, regulatory scrutiny and difficulties raising capital in some markets.

China's competitive advantage lies partly in manufacturing scale and the ability to develop, test and refine physical products through extensive supplier networks. The country's industrial base can support robotics, smart devices and other AI applications that require integration between software and hardware.

The United States, meanwhile, retains major strengths in advanced AI research, computing infrastructure and the financing of leading AI developers. The two countries therefore have different advantages, and their competition is unlikely to be resolved by model performance alone.

For businesses, the immediate priority is often practical: whether a tool can perform a task accurately, securely and at an acceptable cost. For policymakers, the challenge is more complex, involving data security, export controls, supply-chain resilience and technological independence.

The continued use of Chinese AI products by some U.S. companies does not eliminate the strategic rivalry. It demonstrates instead that commercial relationships can persist even when governments are trying to reduce technological dependence. How regulators balance security concerns against cost and innovation will help shape the future of the global AI market.