Stanford University through the Human-Centered Artificial Intelligence Institute (HAI) has published its ninth annual AI Index 2026 report, one of the most authoritative global overviews of the artificial intelligence market.
The study covers hundreds of indicators: investment, model quality, labor market, scientific publications, patents, infrastructure, and adoption of AI in the economy.
The report’s key finding is that China has nearly closed the gap with the US in terms of AI model quality. While in 2023 the difference between the two countries’ best models ranged from 17.5% to 31.6%, by March 2026 it had shrunk to 2.7%.
The authors of the study note that during 2025, U.S. and Chinese models repeatedly swapped places in the world performance rankings. In February 2025, the Chinese model DeepSeek-R1 temporarily reached the level of the best American systems, and by March 2026, the leadership was maintained by the Anthropic model with a minimal advantage.
China catches up with the US with significantly less investment
One of the most discussed AI Index findings was a comparison of AI industry funding levels.
According to the report, U.S. private investment in artificial intelligence in 2025 reached $285.9 billion, while Chinese investment amounted to $12.4 billion. Thus, the U.S. market has invested about 23 times more than China in AI development.
Despite this, Chinese companies were able to almost equal the US in terms of model quality. The authors of the study attribute this to the high efficiency of Chinese developers, the active implementation of AI in industry and large-scale government support for the technology sector, SynapNews explains.
At the same time, the United States retains its leadership in the number of advanced models, data centers and the volume of computing infrastructure. According to Stanford HAI, there are 5,427 data centers in the U.S. – more than ten times more than in any other country.
China leads in patents and industrialization
The report shows that China already dominates a number of structural indicators of the AI economy. In particular:
– In terms of the number of scientific publications and citations in the AI field;
– Chinese companies account for about 69.7% of global AI patent applications;
– China is far ahead of the US in the adoption of industrial robots.
Stanford HAI notes that the US remains the center of development of the most expensive and computationally complex models, but global competition in AI has become significantly more dense than two years ago.
AI is becoming the infrastructure of the global economy
The study also records a dramatic acceleration in the adoption of generative AI in business and education.
According to AI Index:
– 88% of organizations are already using AI tools;
– four out of five university students regularly work with generative AI;
– global investment in AI, including M&A and corporate investments, exceeds $581 billion in 2025.
At the same time, Stanford HAI warns of the growing risks of industry opacity. More than 90% of advanced AI models are created by private companies, and most developers stop disclosing data about the training process of the systems and the datasets used.
What the report’s findings mean
AI Index 2026 shows that the global AI market is moving from a phase of dominance by a few US companies to a more competitive model where China is becoming a full-fledged technological rival to the US.
For businesses, this means increased global competition in cloud platforms, chips, enterprise AI and digital infrastructure. For states, it means the growing importance of technological sovereignty, AI regulation and access to computing resources.
The report also shows that capital efficiency is becoming as important as the absolute amount of investment. Chinese developers have managed to significantly narrow the technological gap with the US even in the face of restrictions on the supply of advanced chips and much less funding, TNW notes.
