U.S. Report Warns China’s Open AI Expansion Threatens the American Edge

By Tahani Elghazaly
Published March 24, 2026

Latest news on WhatsApp أجدد الأخبار على واتساب

A new U.S. report warns that China is no longer challenging the United States in artificial intelligence only through advanced chips, but through the rapid expansion of lower-cost open models that can spread more widely and evolve faster. The U.S.-China Economic and Security Review Commission, a congressional advisory body, said this approach is giving Beijing a compounding advantage that could reshape the global balance of power in AI.

According to the report, major Chinese firms and labs have pushed aggressively into open models while lowering usage costs, helping their systems gain traction around the world at a faster pace. The Commission said Alibaba’s Qwen ecosystem had become the largest on Hugging Face by number of derivatives, with more than 100,000 derived models at the time of publication, highlighting the scale of China’s growing footprint in open AI.

The report argues that China’s strength goes beyond software. It is increasingly tied to the integration of AI into manufacturing, logistics, and robotics, creating a steady flow of real-world data that can improve models over time. It said that advantage could become even more important as the industry moves beyond chat-based systems toward broader applications such as robotics and autonomous systems.

The Commission also said U.S. export controls on advanced chips have focused mainly on slowing model training, but do not fully address the gains that come from large-scale deployment inside the real economy, where new data is constantly generated through everyday use in industry, research, and services. In that context, the report said open-model ecosystems can create alternative paths to leadership even if U.S. companies remain ahead on some technical benchmarks.

The warning reflects growing concern in Washington that the AI race with China is no longer defined only by funding or semiconductor access, but also by deployment speed, affordability, and the ability of an industrial economy to generate large volumes of valuable real-world data.