Trajectories and Comparative Analysis of Global Countries Dominating AI Publications, 2000-2025
This study analyzes OpenAlex data from 2000 to 2025 to reveal a dramatic shift in global AI research dominance, where China has surpassed the US and EU to become the leading contributor in both publication volume and high-impact output, fundamentally restructuring the international AI landscape.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Artificial intelligence has become a defining force of our time, reshaping everything from how we diagnose diseases to how nations compete for economic power. At the heart of this revolution lies a simple but profound question: who is actually doing the work? For decades, the answer seemed obvious. The United States and Western Europe were the undisputed leaders, producing the vast majority of the scientific papers that form the foundation of new technologies. These academic articles are more than just records of discovery; they are the blueprints for future innovation, signaling where the field is heading and which countries hold the most influence. To understand the future of AI, one must first understand the shifting landscape of who is writing these papers.
A recent study by Jason Hung, a research fellow at the Internet Society, takes a long, hard look at this landscape by tracking the publication habits of the world's leading nations from the year 2000 through 2025. The researcher did not simply count how many papers each country wrote, a method that can be misleading when scientists from different nations work together on the same project. Instead, the study used a careful counting method that splits credit for co-authored papers among the countries involved, ensuring a fair picture of each nation's true contribution. By analyzing millions of records from a massive, open database of scientific literature, the study maps out exactly how the balance of power has tipped over the last quarter-century.
The story the data tells is one of dramatic transformation. In the year 2000, the United States and the European Union together produced more than half of the world's AI research, holding a combined share of over 57 percent. China, at that time, contributed less than 5 percent. By 2025, the situation has flipped entirely. The combined share of the US and the EU has plummeted to less than 25 percent. In their place, China has surged to become the single most dominant contributor, accounting for nearly 36 percent of all global AI publications. This is not a slow drift but a rapid, policy-driven ascent. While the US and Europe have seen their relative shares cut by more than half, China's output has grown exponentially, moving from a minor player to the central engine of global AI research.
The shift is even more striking when looking at the quality of the work, not just the quantity. A common assumption in the scientific community has been that while China might produce more papers, the West still holds the advantage in high-impact, influential research. The study challenges this belief directly. By examining which papers are cited most frequently by other scientists—a standard measure of influence—the research finds that China now leads not only in volume but also in the share of high-impact publications. In 2025, China's share of the most influential papers was significantly higher than its share of total papers, suggesting its research output is disproportionately concentrated in high-impact work relative to its volume. However, the study explicitly notes that citation-based measures proxy influence rather than quality, and this finding alone does not establish that Chinese AI research is qualitatively superior. Meanwhile, the average number of citations per paper for the United States has dropped sharply, falling from about 70 citations per paper in 2000 to roughly 15 by 2025. China's average citations have risen steadily, surpassing the US around 2018 or 2019.
This new reality is not just about China rising; it is also about the fragmentation of the West. The study notes that the European Union, once a unified powerhouse, has seen its collective share decline steadily, a trend exacerbated by the departure of the United Kingdom and the lack of a cohesive, centralized strategy compared to other regions. In the United States, a significant portion of top AI talent has migrated from universities to private technology companies, meaning their groundbreaking work often appears in corporate labs rather than in the public academic records the study tracks. This migration helps explain why the US share of academic papers has shrunk even as the country remains commercially dominant.
The data also reveals that the global AI research landscape is becoming more concentrated, with a smaller number of players controlling a larger share of the output. The study identifies a critical turning point around 2012, coinciding with major breakthroughs in deep learning and a surge in China's national investment in AI. Since then, the gap between the dominant players and the rest of the world has widened, rather than narrowed. While India has also emerged as a significant player, growing its share to nearly 10 percent, the overall trend points toward a multipolar world where Asian nations, led by China, are setting the pace. The study suggests that the historical dominance of Western powers is not just a temporary fluctuation but a structural change, driven by centralized national strategies in Asia and the dispersal of research efforts in the West.
Ultimately, this research paints a picture of a field that has fundamentally reorganized itself. The era where the US and Europe quietly set the agenda for artificial intelligence is giving way to a new dynamic where China is the primary driver of both the volume and the influence of scientific discovery. This shift carries deep implications for the future, as the country that produces the most foundational knowledge is often the one that shapes the standards and technologies of tomorrow. The study concludes that the global race for AI is no longer a contest between a few established giants, but a complex, multipolar competition rooted in the rapidly evolving research ecosystems of Asia.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.