On March 30, 2026, the Bayuegua Institute of Science and Technology Innovation, a sci-tech innovation think tank service institution, officially released the Global AI Enterprise Technology Innovation Index Report 2026, revealing a Sino-US dual-core leadership and prominent industrial agglomeration effect.

Relying on global patent big data and authoritative industrial information, the report takes core micro-enterprises in the global AI field as the research object for the first time, presenting a comprehensive picture of the innovation strength and industrial landscape of global AI enterprises.

With reference to the Guidelines for the Construction of a Comprehensive Standardization System for the National Artificial Intelligence Industry (2024 Edition), the report has constructed a "four tiers and one category" research framework, namely the infrastructure layer, framework layer, model layer, application layer plus comprehensive large enterprises, covering the entire AI industrial chain in an all-round way.

On this basis, it has built a technology innovation evaluation system consisting of 4 first-level indicators — innovation output, innovation quality and value, operational capacity and commercialization, and capitalization and global layout — and 20 second-level indicators.

The report shows that the global AI industry has formed a highly concentrated regional pattern. Among the 100 benchmark enterprises counted in this report, 51 are from China and 37 from the United States, accounting for 88% of the total worldwide. The overall distribution features a Sino-US dual-core, with Europe as a supplementary force and other regions as additional supplements.

At the same time, the spatial agglomeration effect of global AI enterprises is extremely prominent, with 51% of the enterprises concentrated in four core sci-tech cities: San Francisco, Beijing, Shanghai and Shenzhen.

Among them, Chinese AI enterprises are highly concentrated in leading cities — 26 enterprises are based in Beijing, accounting for 51% of the domestic listed enterprises; in North America, the San Francisco Bay Area serves as the core industrial hub, gathering 24 enterprises, accounting for 64.9% of the US listed enterprises, and is the global core cradle of AI innovation.

Notably, this pattern is not a simple quantitative comparison. The report points out that China and the United States have formed a distinct "complementary competition" in the AI industrial chain.

Beyond the macro pattern, the report focuses closely on the entire AI industrial chain, conducting in-depth analysis of the innovation strength of enterprises at different levels, and clearly presenting the competitive characteristics, strengths and weaknesses of all links in the global AI industrial chain.

It shows that the 100 listed enterprises as a whole present an industrial chain distribution pattern dominated by the application layer, with equal emphasis on the infrastructure and framework layers, and supplemented by the model layer and comprehensive large enterprises.

Among them, China holds a quantitative advantage in the infrastructure layer, model layer and application layer, especially in the implementation of AI application scenarios; the United States leads by a wide margin in the framework layer, dominating high-barrier fields such as open-source AI ecosystems and core development tools, and holds absolute discourse power in the high-end chip market of the infrastructure layer.

This differentiated industrial layout has jointly shaped the current competitive pattern of the global AI industry.

As the core foundation of the AI industry, 20 enterprises are listed in the infrastructure layer, including Nvidia, Intel, AMD, Inspur, Qualcomm, Sugon, and Cambricon. China has a quantitative advantage in the number of enterprises, but there is considerable room for improvement in its globalization system.

The United States maintains an absolute leading position in the framework layer, with 11 out of the 20 listed enterprises from the US, which dominates high-barrier fields such as open-source AI platforms and data services.

A total of 10 enterprises are listed in the model layer, forming a competitive pattern where Chinese enterprises take the lead in quantity and US enterprises lead in technology.

The application layer has the largest number of listed enterprises, with 40 in total, serving as the core link for the commercialization of the AI industry. Among them, 21 are from China, boasting outstanding capabilities in scenario implementation.

10 comprehensive large enterprises are also listed, showing distinct echelon differences in patent filings and global layout. Among them, Samsung ranks first in the world with 2,073 PCT patent filings, and Huawei takes the lead among domestic enterprises in China.