The world of artificial intelligence is currently experiencing a surprising turning point. For a long time, ChatGPT, developed by the US company OpenAI, was seen as a symbol of Western dominance in the field of large language models. But at the beginning of 2025, the Chinese start-up DeepSeek caused a real sensation with its AI software “R1”. DeepSeek developed a language model that it claims is just as powerful as the leading US models, but with significantly less energy, computing power and, above all, costs. DeepSeek reportedly only invested around six million euros in the model, while the training and development of ChatGPT is said to have cost more than ten times as much. The success not only shook the market, but also had direct economic consequences: The shares of US chip giant Nvidia lost around 17% of their value in just one day after the news broke – a record loss of around USD 590 billion.
Translated with DeepL.com (free version)

AI as a key technology

AI has long since become a key geopolitical technology, and the race between the USA and China has gained new momentum thanks to DeepSeek. Since 2017, China has been pursuing the declared strategy of becoming the leading AI nation by 2030. While the USA has traditionally been a pioneer in large-scale language models, Chinese companies are rapidly catching up. In addition to DeepSeek, giants such as Alibaba, Tencent and Bytedance are increasingly publishing models that can compete with Western standards. These models are often made available as open source, which further increases their global appeal.

China’s race to catch up

China’s strength is also growing visibly in the area of image and video models. The first powerful AI video model, CogVideo, came from China as early as 2022 and was months ahead of the competition. Chinese companies are benefiting from large amounts of visual data collected by popular platforms such as TikTok. Alibaba’s Chinese model “Wan 2.1” is considered particularly advanced and has been made freely accessible worldwide, further increasing the pressure on Western competitors.

Data – the new oil

Data is considered the “new oil” for AI development. The USA still has significant advantages here thanks to its large tech companies such as Amazon, Google and Microsoft as well as specialized start-ups such as Scale AI. The market for AI training data in the USA already has a volume of around 580.5 million US dollars. Although China is catching up quickly, it suffers from restrictions on data quality and accessibility due to government regulation and censorship.

Talent factory vs. talent magnet

A decisive factor in the AI competition is the training of AI talent. According to Gartner, 47% of the world’s top AI researchers come from China, while only 18% come from the USA. However, many Chinese researchers are leaving the country for the USA – a major risk for China’s long-term goals. But China is taking countermeasures: The number of Chinese universities offering AI degree programs rose to 537 in 2024. According to Feiyu Xu, former head of AI at SAP, China’s talent development starts very early: “In China, children learn AI as early as kindergarten.” The education system specifically promotes talent according to the country’s needs in order to keep up internationally.

Stronger together

Despite the intense competition, the USA and China are also working closely together. Zachary Arnold from the Emerging Technology Observatory emphasizes that international cooperation is crucial for progress:

“AI is a highly collaborative field. There are diverse AI researchers and engineers all over the world. International cooperation significantly accelerates development.”

According to recent studies, joint research projects between the two countries have a particularly high impact. Between 2014 and 2024, the USA and China were the most frequent partners in international AI collaborations. Such collaborations are particularly valuable because they enable AI models to be tested and further developed in different cultural, social and economic contexts. This results in more robust and versatile solutions. Despite geopolitical tensions and national interests, collaboration is widespread and productive, particularly in areas such as computer vision or the development of new AI algorithms.

How is Europe doing?

Europe’s position appears more reserved at first, but the continent certainly has the potential to pursue its own strategy. According to André Retterath from Earlybird VC, Europe has the advantage of highly specialized expertise and high-quality data sets from SMEs. Europe’s strength could lie in combining industry-specific expertise with existing AI models. Digital Minister Volker Wissing is calling for better conditions for venture capitalists and fewer regulatory hurdles. The EU’s comprehensive AI regulation could act as both a brake on innovation and a quality feature.

Prospects: Who will win the race?

The current AI competition shows that China is massively catching up and seriously challenging the USA. The example of DeepSeek makes it clear that the previous dominance of Western companies is not guaranteed. Although the US remains ahead in terms of transformative innovation and AI infrastructure, China’s advances in industrial applications, data utilization and talent development are putting increasing pressure on the West. Europe’s opportunity could lie in leveraging its strengths to establish itself as a center for specialized, high-value AI applications and solutions. The outcome of the competition remains to be seen, but one thing is clear: AI has long since become not just a technological innovation, but the central arena of geopolitical power struggles.

Cover picture © stock.adobe/Танюша Коновал

Sources:
AlShebli, Bedoor; Memon, Shahan Ali; Evans, James A.; Rahwan, Talal (2024): „China and the U.S. produce more impactful AI research when collaborating together“, in: Scientific Reports, 14, Artikelnummer: 28576. Online verfügbar unter: https://www.nature.com/articles/s41598-024-79863-5 (Access: 19.03.2025).

Amal, Khadija (2025): „US and China lead global AI collaboration as research partnerships grow“, in: Rest of World. Online verfügbar unter: https://restofworld.org/2025/us-china-lead-global-ai-collaboration/ (Access: 19.03.2025).

Kreimeier, Nils (2025): „Kampf um KI: Wie Europa gegen USA und China bestehen kann“, in: Capital.de. Online verfügbar unter: https://www.capital.de/wirtschaft-politik/kampf-um-ki–wie-europa-gegen-usa-und-china-bestehen-kann-35530576.html (Access: 19.03.2025).

Deutschlandfunk (2025): „Schlüsseltechnologie im Wettstreit zwischen China, USA und Europa“. Online verfügbar unter: https://www.deutschlandfunk.de/schluesseltechnologie-im-wettstreit-zwischen-china-usa-und-europa-100.html (Access: 19.03.2025).