Despite tensions, U.S.-China AI research collaborations are alive and well
- Writer: info@mps-asia.com at
- Tech news
Over the past 10 years, the U.S. and China have been the most frequent partners in AI research.
The field of artificial intelligence has long been dominated by the U.S. and Europe. Recently, however, AI research has become more collaborative across countries outside of the West, according to a Rest of World analysis. The findings show that over the past 10 years, AI researchers in non-Western countries are participating in more transnational research collaborations.
Rest of World’s analysis used data collected by the Emerging Technology Observatory at Georgetown University in Washington. Their Country Activity Tracker tool measures various indicators of AI innovation by country, including patents, investments, and research papers. The tool pulls from six scholarly research platforms to maintain a database of over 260 million research papers.
Rest of World used this data to identify which non-Western countries have produced the most collaborative AI research papers in the past 10 years, and which other countries were their most common collaborators. For a country to have published a research paper, at least one co-author from an organization in that country must be involved.
“AI is a very highly collaborative field. There [are] varied AI researchers and engineers all over the planet,” Zachary Arnold, an analyst at the Emerging Technology Observatory, told Rest of World. “By being engaged with each other in collaboration, I think most people would say the field’s accelerated.”
Studies have shown that cross-border collaboration in research, involving people with diverse backgrounds and values, is important across all scientific disciplines to ensure that a variety of perspectives and social contexts are reflected in the findings. This is equally true for the development of AI.
“In applied AI research, one of the very common ways promising AI tools or approaches tend to break down is you move them even slightly out of the context of the particular lab or particular society that they were developed for. Things go sideways very quickly in ways you didn’t predict,” Arnold said. “Having cross-border collaborations where you can test in different contexts could be very important.”
According to the data, over the past 10 years, the U.S. and China have been the most frequent partners in AI research. More recently, China’s collaboration with other countries like Singapore and Saudi Arabia is also steadily rising.
Rest of World found that the top 10 non-Western countries by number of collaborative AI research papers published between 2014 and 2024 are China, India, Japan, Singapore, South Korea, Saudi Arabia, Pakistan, Malaysia, Brazil, and Taiwan.
Data for 2023 and 2024 is incomplete due to the lag between the start of new research and its publication. However, all of these countries have already published more AI research papers with transnational collaboration so far in 2023 than in 2022, suggesting the upward trend will continue.
Countries that teamed up to do AI research most frequently partnered with China, the U.S., or a European country. Saudi Arabia was an outlier — its top collaborators were Pakistan, India, and Egypt. Malaysia’s top collaborators were China, India, Saudi Arabia, and Indonesia.
Many of these research collaborations are related to the field of computer vision, which concerns training computers to interpret information from photos and videos for use in facial recognition and autonomous vehicles, among other practical applications. In these papers, researchers are tackling new ways to use this technology — a vital first step to drive forward AI innovation, Saaidal Razzali Azzuhri, a senior lecturer at the University of Malaya in Kuala Lumpur, told Rest of World.
A current research collaboration between the China Academy of Information and Communications Technology, University of Malaya, and Malaysian blockchain company Zetrix is one example of cross-border research that bridges the public and private sectors. The project is developing ways to utilize blockchain and AI technologies to bolster trade between China and Malaysia.
Governments play an important role in supporting cross-border research, said Azzuhri, the project’s lead researcher. “When you want to do this kind of research, you must have government support.”
Azzuhri hopes there will be a rise in Malaysia-U.S. collaboration in the future, since big U.S. tech companies including Microsoft and Amazon have set up data centers in the country. He thinks building infrastructure like large-scale data centers can help attract more researchers interested in collaborations, as well as investment.
“Maybe there are some future problems that we don’t know yet,” Azzuhri said. “You have to do cross-border research to solve [cross-border] problems that arise.”