The U.S.-China AI rivalry is shifting from a focus on model performance to include access to computing infrastructure, financing costs, and the ability to translate AI advances into real productivity and revenue.

At the September 24 summit, there was no decisive agreement on AI limits or a formal joint safety framework, and proposals for incident notification between the two nations remain unfinalized.

Overall, the summit did not change the trajectory of the rivalry; future developments will depend on computing capacity, financing conditions, industrial deployment, and actual productivity gains rather than model performance alone.

Alibaba unveiled the Zhenwu V900 AI accelerator and aims to expand global data-center capacity to more than 20 gigawatts by 2032, while Huawei remains a major supplier of China’s AI infrastructure, though Chinese models reportedly lag frontier U.S. models in some areas.

China is boosting its AI capabilities by building domestic processors, cloud infrastructure, and data centers in response to restrictions on advanced chips, with firms like Huawei and Alibaba accelerating investment in chips, cloud services, and AI models.

Rising long-term U.S. interest rates above 5% are raising the cost of financing AI infrastructure, data centers, and power networks, shaping project economics and investment decisions.

China’s strategy emphasizes building domestic computing capacity and broad AI deployment across industries, while the U.S. remains strong in frontier models, semiconductors, cloud platforms, and capital markets.