AI from Germany: Europe’s smart alternative

“We finally need to stop warming up and just get on with the race.” These were the views of Susanne Dehmel, board member of German digital association Bitkom, expressed recently regarding a development that has caught the attention of AI developers worldwide: Chinese company DeepSeek released a so-called basic model that manages with significantly fewer resources than the current leading systems from the USA. Such models usually require vast amounts of data and to be trained over a long time. “DeepSeek shows that the AI market is still a lot more dynamic than assumed, that neither the winners or losers are yet determined – and that the USA is still far from enjoying a monopoly”, according to digital expert Dehmel. This is seen as good news in the discussion surrounding digital sovereignty in Germany and Europe. 

Smaller and more efficient 

How did DeepSeek do it? Kristian Kersting feels that the surprise success from China is an indication that size alone is not a prerequisite for top-quality results. Kersting is Professor of Artificial Intelligence and Machine Learning in the Department of Computer Science at TU Darmstadt and one of Germany’s leading AI experts. He believes: “You need to break the tasks down into smaller elements and adapt your AI to their level of difficulty. The arithmetic task 2 plus 2 doesn’t of course require as much capacity as the issue of solving the climate crisis.” 

Partnerships with local companies 

This also shines a different light on what at first sight looks like sobering news. It was announced in September 2024 that Aleph Alpha, probably the best known German AI startup and one of the hopeful contenders in the race for the best large basic models, was withdrawing from this power struggle. The company intends instead to focus on smaller models for specific applications, according to Aleph Alpha CEO Jonas Andrulis. This very approach could pay off in the medium term. French company Mistral, alongside Aleph Alpha one of the hopeful AI contenders at a European level, is also following this path. And with considerable success: Mistral’s CEO Arthur Mensch believes that increasing numbers of European enterprises want to form partnerships with local companies. Aleph Alpha is already successfully cooperating with German SMEs. “It’s ultimately a question of monetising such models – in other words turning them into tangible applications. And this is a path that I believe we are following more consistently in Germany and Europe than elsewhere”, says AI Professor Kersting. 

200 billion euros for AI in Europe 

Since the AI Summit in Paris in February 2025, significantly more capital has been made available for AI development in Germany and Europe. This is urgently needed for the development of AI models. The EU wants in future to invest up to 200 billion euros in AI development on the continent.  

Leading AI startups from Germany 

This could also benefit a wide range of German AI companies that are already internationally competitive. Celonis, a spin-off from the Technical University of München, is currently valued at tens of billions. This spin-off helps companies analyse and optimise their business processes. The startup Black Forest Labs is a global leader in the AI generation of images and is building on the research results of Björn Ommer, an internationally recognised leading AI researcher from LMU Munich. And Munich defence contractor Helsing is seen as one of Europe’s fastest-growing companies in defence technology. It is valued at five billion euros. Perhaps the race has already started after all.