• The US, UK, and EU are still in the AI race, but cracks are showing.
• China leads in output, patents, and talent, while the West leans on global ties.

AI has become a matter of national interest for governments worldwide, with some treating the technology as if it were like energy or defence. While China gains attention for rapid growth in the space, a closer look shows that the West – namely the US, UK, and EU – is still a major player in AI research, even if the balance is starting to shift.

A recent report by research analytics company Digital Science tracks global AI research trends from 2000 to 2024. Based on data from the Dimensions research database, the report compares research volume, collaboration, talent movement, and innovation output in different regions. Figures show that the West remains influential in AI, but also points to areas where it’s starting to lose ground.

The US holds ground but faces pressure

The United States still plays a big role in AI development. It maintains the strongest ecosystem for startups and company-led research. While the country no longer leads in the number of academic publications – that title now belongs to China – it continues to drive commercial applications of AI, especially through firms like OpenAI, Google, and Anthropic.

However, much of that research happens behind closed doors. The report notes a drop in publicly-available AI research from US firms. As companies focus more on private development, the visibility and academic impact of US-based AI research may be harder to track.

On paper, the US remains the second-largest contributor to AI research by volume. It also plays a central role in international collaboration. But even here, things are changing. China has become the US’s top AI research partner, and that relationship has continued despite rising political tension. Meanwhile, the US is more dependent on the partnerships than China, which publishes most of its research without Western involvement.

Patent filings show another area of concern to Western observers. While the US has a strong record of commercialising research, China now files about ten times more AI-related patents. That suggests the US may be publishing quality work but not always turning it into protected innovation.

UK punches above its weight

Despite its smaller size, the UK continues to perform well in AI research impact. It ranks fourth globally in attention-per-output, which measures how often a country’s research is cited relative to how much it publishes: UK research is being noticed, even if it isn’t produced in high volumes.

The UK also has strong AI hubs in London, Cambridge, and Edinburgh, with several institutions consistently publishing high-impact work. But it leans heavily on international partners to maintain that status. China has become its top AI research collaborator – accounting for over 25% of the UK’s collaborative output in 2024, more than the UK shares with the US or EU.

The country’s total number of AI researchers, about 3,000, is modest compared to China or the EU but the UK continues to attract talent and attention. The challenge is keeping the momentum going as funding shifts and global competition increases.

EU shows internal strength, external gaps

The EU-27 collectively produces a large volume of AI research, second only to China. Many of its member countries – Germany, Italy, Spain, and the Netherlands – have active research communities. EU institutions also benefit from regional programmes that support collaboration inside individual countries’ borders. That internal cooperation is a strong point.

Where the EU falls short is in visibility and conversion. Its research doesn’t attract as much global attention, and fewer projects result in patents or market-ready solutions. Only a small portion of its research output is internationally cited at high levels. The report suggests that while EU countries are producing a lot of work, they may struggle to get it noticed or applied.

External collaboration is another weak spot. Compared to the US and UK, the EU collaborates with fewer countries outside the region. This limits exposure and reduces the chance for shared development with global leaders in AI.

The EU is also fragmentated. While there is internal cooperation, there’s less clarity on a unified AI strategy in the bloc. That leaves some research institutions without the scale or coordination needed to compete with larger players.

Talent is moving, and the West is losing

One of the more striking data points in the report involves international moves by researchers. In the past, Western countries were top destinations for AI talent but that pattern is changing.

China now has a net gain in the number of AI researchers from the US, UK, and EU. In other words, more researchers are moving to China than leaving it. Meanwhile, the West is still drawing in international talent, but it’s losing some of its own experts in return.

The US remains the largest source and destination for AI researchers globally, but its net balance has shrunk. The UK and EU are also sending more talent abroad than they’re getting back. The raises questions about how long the West can maintain its research base without stronger local investment and retention.

Institutions in the West still matter

While the report focuses on national output, it also highlights standout institutions. The US is home to several high-output AI institutions, including top universities and research labs. The UK has 19 institutions that each published over 50 AI papers in 2024. In the EU, that number stands at 54. For comparison, China has 156 such institutions.

The disparity shows that while quality exists in the West, it’s less evenly spread. China’s broader base gives it more resilience and reach. The US and EU have high-performing research centres, but they tend to be more concentrated.

Collaboration patterns are shifting

International collaboration has long been a strength of Western research. The report confirms that trend but adds an important caveat: while the West is collaborating more, it’s also growing more reliant on China.

In 2024, China was the top AI collaborator for the US, UK, and EU. But China does not rely on those same countries to the same extent. Just 4% of China’s AI research included collaborators from the West. That number was 25% for the UK and about 10-12% for the US and EU.

The imbalance suggests that China has positioned itself as a connector – but one that can also work alone. The West, in contrast, depends more on external ties to maintain research output and influence.

Looking ahead

The West is still in the AI race – but it may need to rethink its strategy. While the US, UK, and EU have strong research histories, high-impact institutions, and skilled talent, those advantages are being tested.

As China expands its AI base, builds long-term talent, and dominates in patents, Western nations may need to focus less on volume and more on outcomes. That includes better research-to-product pipelines, more inclusive collaboration networks, and policies that keep talent from drifting elsewhere.

The gap isn’t yet unbridgeable – but the numbers show it’s widening.