News Story

Stanford releases the 2025 AI Index, with a little help from KMi

Nancy Pontika, Wednesday 9 Apr 2025

The Stanford Institute for Human-Centered AI published its 2025 AI Index report this week, providing a comprehensive look at the global state of Artificial Intelligence (AI). The report, now in its eighth edition, tracks and visualises data on technical performance, economic impact, education, policy, and responsible AI to offer an empirical foundation for understanding the field’s rapid evolution.

The AI Index is a globally recognised source for credible and authoritative data and insights on AI. It has been cited in major news outlets like The New York Times, Bloomberg, and The Guardian, and referenced by policymakers in the United States, the United Kingdom, and the European Union. The AI Index’s growing significance is reflected in the increased size, scale, and scope of this year’s edition, which surpasses all previous versions.

The compilation of the AI Index relied on help from KMi. In particular, Dr. Angelo Salatino, a key member of KMi’s Scholarly Knowledge Modelling team (SKM), was awarded $10,000 from Stanford University to support the development of the AI Index. Given his extensive expertise in the field, over the past 10 years, he supported and supervised the Stanford team in developing various statistics around papers in AI. These included the number of publications and citations in the last 10 years, and how these compare with the whole Computer Science field. These publications were then categorised by country and region (e.g., North America, Europe); sector (e.g., academia, industry, government, non-profit); and AI subdiscipline (e.g., Machine Learning, Computer Vision, Natural Language Processing).

Part of these activities also included updating KMi’s Computer Science Ontology (CSO) with the latest trends in AI and further developing the CSO Classifier, the team’s flagship product, with additional features tailored for this project. These two KMi technologies were then used by the Stanford team to categorise nearly 2 million AI research papers, taking advantage of the highly detailed classification of research topics provided by CSO – indeed, CSO provides the most granular and comprehensive taxonomy of research topics in Computer Science currently available.

Interesting insights emerged from this research:

Industry continues to make significant investments in AI and leads in notable AI model development, while academia leads in highly cited research.
China leads in AI research publication totals, while the United States leads in highly influential research.
AI publication totals continue to grow and increasingly dominate computer science.
The United States continues to be the leading source of notable AI models.

To learn more about these insights, you can read the report (available at the end of the page, in the “Related links” section.

On LinkedIn, Angelo mentioned that he is extremely honoured to have been involved in this initiative, and he is grateful to Dr. Loredana Fattorini and Dr. Nestor Maslej, both researchers at Stanford University for inviting him to join this initiative.

The whole SKM team is pleased by the success of the project and they are looking forward to further collaborating with the Stanford team on new exciting projects, including the 2026 AI Index.