No single ranking decides which country leads in AI, but the United States comes first in the two best-documented country indexes: Stanford HAI's Global AI Vibrancy Tool (November 2024 edition, 36 countries) and Oxford Insights' Government AI Readiness Index 2025 (195 governments). Below first place the lists diverge sharply. China placed second in the Stanford tool and eighth in the Oxford index, and Canada placed 12th in the Oxford index. The gap comes from what each index counts, not from disagreement about the data.
Key takeaways
- Stanford's tool ranked the United States first, China second, and the United Kingdom third in November 2024, using 42 AI-specific indicators across 36 countries.
- Oxford Insights' 2025 index ranks governments, not whole economies, and places the United States first, China eighth (up from 23rd in 2024), and Canada 12th.
- Stanford's tool lets users change the weights, so the order of countries is a product of choices the user can alter.
- China leads on AI patenting in Stanford's findings even where it ranks lower overall, which shows how a single indicator can reverse a story.
What each index actually measures
The Stanford Institute for Human-Centered AI launched its Global AI Vibrancy Tool on November 21, 2024. It scores 36 countries on 42 indicators grouped under pillars that include research and development, economy, and infrastructure. The institute says the United States leads in most pillars, while China leads in AI patenting. It also describes the launch as a start and says more data and countries are planned. A 2025 update was posted on November 24, 2025 with data through 2024 and a streamlined indicator set, but the country table was not available in text form on the page, so this article does not quote 2025 positions.
Oxford Insights asks a different question: to what extent a government can use AI for public benefit. The 2025 edition assesses 195 governments and, after a methodology change, looks at governments as buyers, enablers, and regulators of AI. Earlier editions concentrated on AI in public services. The index has run since 2017. Because the questions differ, a country can rank well on one list and modestly on the other without any contradiction.
Two lists side by side
| Question | Stanford HAI vibrancy tool | Oxford Insights readiness index |
|---|---|---|
| Unit of comparison | National AI ecosystem | Government capacity |
| Coverage | 36 countries (2024 edition) | 195 governments (2025 edition) |
| First place | United States | United States |
| China | Second | Eighth, up from 23rd in 2024 |
| United Kingdom | Third | Not reported in the sources checked |
| Canada | Not reported in the sources checked | 12th |
| User control | Weights adjustable | Filters by region and income group |
The Stanford release also reported the United Arab Emirates fifth, France sixth, and South Korea seventh in 2024.
Why China's position moves so much
China's jump from 23rd to eighth in the Oxford index came in an edition with a changed methodology, and Oxford Insights itself argues that the position probably understates China's capabilities because of limited transparency and restricted access to deployment data. The page also notes that China's 2024 AI research output was reported to match the combined publications of the United States, the United Kingdom, and the European Union. In other words, the rank reflects what a surveyor can see, and measurement access varies by country.
How weighting can reorder a list (illustrative example)
The numbers in this section are invented to show the mechanism and are not data from either index. Suppose three countries score out of 100 on two pillars.
| Country | Research score | Infrastructure score | Rank if research counts 70 percent | Rank if infrastructure counts 70 percent |
|---|---|---|---|---|
| A | 90 | 40 | First (75) | Second (55) |
| B | 60 | 80 | Second (66) | First (74) |
| C | 50 | 50 | Third (50) | Third (50) |
Country A and country B swap places without any change to their underlying scores. Stanford's adjustable weights make the same point on a real dataset, and any published ranking embeds a weighting even when the user cannot see it.
Questions to ask before quoting a ranking
- Does the index rank governments, companies, researchers, or whole economies?
- How many countries are included, and which are missing?
- Which year does the data describe, and which year was it published?
- Are the weights fixed, and can they be changed?
- Is the source an academic institute, a consultancy, or a media outlet selling a list?
FAQ
Which country is first in AI?
The United States, in both the Stanford HAI tool (2024 edition) and the Oxford Insights readiness index (2025 edition). No ranking covers every dimension of AI, so first place on one list does not carry over to the others.
Where does Canada rank?
Oxford Insights places Canada 12th in its 2025 government readiness index. A Canadian position in the Stanford tool was not found in the sources checked.
Is China behind the United States in AI?
It depends on the measure. China ranked second in Stanford's tool and leads in AI patenting there, but eighth in Oxford's government index, where limited transparency affects the score.
Are these rankings official?
No. They are published by a university institute and a research consultancy, not by a government or international body, and their methods can change between editions.
Related reading
- The Global Partnership on AI
- Where AI education is concentrated
- AI in research and discovery
- DeepMind research at ICLR 2023
More in AI Tools and Industry.
Last verified: October 2026
Sources
- Global AI Power Rankings: Stanford HAI Tool Ranks 36 Countries in AI, Stanford HAI, November 21, 2024
- The Global AI Vibrancy Tool 2025, Stanford HAI, November 24, 2025
- Government AI Readiness Index 2025, Oxford Insights, 2025