European AI index
European AI index
Europe is striving to counterbalance American and Chinese AI dominance. But the road is long and paved by structural, financial and technological challenges.
While 73 percent of all American companies use artificial intelligence in everyday life and 83 percent in China, companies in the countries of the European Union are even more cautious.
Especially in Germany, as the largest economy in the European Union, the hoped-for productivity boost from the use of the technology is still missing. According to a study by the Boston Consulting Group (https://www.bcg.com/press/15january2025-zukunftstechnologie-ki-2025-trifft-weltweite-dynamik-auf-deutsche-zuruckhaltung) to regulatory challenges, data protection concerns and fears of loss of control. In addition, many still lack an understanding of the technology and AI implementations such as software introductions are being made, which in turn leads to poor results and disillusionment in management.
For example, many companies only allow their employees to use a selected AI tool that suggests alleged data security. However, excellent LLMs can now be operated locally and dispel data protection concerns. This requires more transparency and independent clarification so that regulations based on obscured sources of information do not lead to a lasting competitive disadvantage and irreversible dependencies.
Although investments in start-ups are increasing slightly, they are still at a manageable level. In addition, European start-ups are also often dependent on American venture capital if they want to grow — or, as in the case of the European flagship AI start-up Mistral AI, enter into close cooperation with already leading AI companies such as Nvidia. Europe still lacks an ecosystem in which young start-ups are adequately financed and promoted.
Start-ups have the best financing opportunities in England, where 17 billion euros are invested annually in young tech and digital companies. France and Germany are in second place with just under 8 billion euros each. By way of comparison, American start-ups were financed with 128.8 billion dollars in 2024.
1.3 billion euros were invested specifically in AI start-ups in France, 910 million in Germany and only 318 million euros in England. All other countries in the EU do not publish reliable figures on this.
Regardless of all the current reluctance of European companies, artificial intelligence will become the game-changing technology of the coming decades. Even though users in general have yet to learn how to use generative AI efficiently, the technology will become deeply rooted in companies and optimize and automate numerous processes.
Accordingly, the digital infrastructure in Europe must be significantly and rapidly expanded. Around 16 percent of the world's hyperscale data centers are currently located in European countries and produce just under 10 GW of computing power with a power consumption of 62 TWh per year, which accounts for around 2 percent of the total electricity consumption of the European Union. By 2030, the expected computing load will be 35 GW and power consumption will rise to 150 TWh.
Building this infrastructure requires not only between 250 and 300 billion euros, but also building land, skilled workers, moderate electricity prices and nationwide fibre-optic expansion.
60 percent of all hyperscale data centers are currently operated by three tech companies: Amazon (AWS), Microsoft and Google. If Europe wants data sovereignty, then future investments must come out of the EU as a lead.
Europe is playing one of the leading roles in the area of future AI technologies. For example, Germany is in second place after China in the field of quantum computer research.
After twenty years of research, the German company Carl Zeiss has achieved a revolutionary breakthrough in chip manufacturing. Under the catchy, catchy name “High Numerical Aperture Extreme Ultraviolet Lithography” (short: High-NA-EUV), Zeiss and its Dutch partner ASML (the only company in the world that is able to produce such precise chip production machines) have succeeded in producing a series production that is already being used by Intel and will result in 100 billion transistors on a chip becoming one trillion transistors in the future.
This technological quantum leap not only makes future AI chips extremely powerful, but also reduces the power consumption of data centers.
Conclusion: Europe is making great progress in the development and use of artificial intelligence. European companies and institutions are leading the way, particularly in research and development of future basic technologies for AI. However, companies must not rely on the chicken-and-egg principle when using artificial intelligence and wait for the European AI infrastructure to catch up with the American one, but must now expand their efforts to integrate artificial intelligence broadly and comprehensively into their processes and workflows.
