The artificial intelligence (AI) landscape in Europe is about to undergo a momentous change.
Last February 2, 2025 not only marked the beginning of the application of the much-debated AI Act, a regulation that imposes new obligations and sanctions on those who use AI in European territory, but also coincides with the announcement of the launch of OpenEuroLLM.
This ambitious project aims to revolutionize the sector through the development of open, multilingual LLM models aligned with European values, demonstrating that innovation can and should coexist with regulation.
A pioneering project in the heart of Europe
OpenEuroLLM was born in response to the need for Europe to regain leadership in the field of artificial intelligence.
With a total investment of 37.4 million euros – of which 20.6 million come from the Digital Europe Program –, it is positioned as the first coordinated effort to create a family of open language models (LLM) capable of covering all the official and future languages of the European Union.
This multilingual approach is key for a region as diverse as Europe, where linguistic plurality has traditionally been a challenge for predominant AI models, developed mostly for English or other majority languages.
The European Commission has recognized the relevance of this initiative by awarding it the prestigious STEP Seal, a mark of excellence within the Strategic Technologies for Europe (STEP) platform.
Regulation and innovation in parallel?
The coincidence of the OpenEuroLLM announcement with the start of the AI Act is no coincidence.
While the new regulations impose obligations on AI providers and users such as guaranteeing an adequate level of AI literacy, preventing uses that can manipulate or exploit people and establishing internal governance mechanisms, OpenEuroLLM stands as a commitment to innovation.
The AI Act seeks to mitigate the risks associated with the use of artificial intelligence through sanctions that can reach up to 35 million euros or 7% of global turnover in cases of prohibited practices.
However, many critics believe that an overly regulatory approach could stifle innovation.
In this context, OpenEuroLLM is presented as an alternative that, without avoiding the requirements of regulations, promotes creativity and international competitiveness.
An open model for a diverse Europe
With 27 member states and more than 30 languages, the continent needs solutions that go beyond large-scale models designed for homogeneous markets.
OpenEuroLLM aims to overcome this barrier by developing a model capable of operating in all official languages and, in the future, those that join the EU.
This feature not only represents a significant technical advance, but also reaffirms the initiative’s commitment to fundamental values such as inclusion, cultural diversity and digital autonomy.
Furthermore, the possibility for companies and public administrations to customize these models according to their needs opens a range of opportunities in sectors as varied as education, health, public administration and commerce.
Collaboration and high-performance resources
The success of OpenEuroLLM also lies in the collaboration of various key players in the European AI ecosystem. More than 20 research institutions, technology companies and supercomputing centers have joined together to promote this project.
Among them are Aleph Alpha from Germany, CSC from Finland and Lights On from France, which provide their experience and technical capacity.
The project is led by Jan Hajič, professor of computational linguistics at Charles University in Prague, and Peter Sarlin, co-founder of Silo AI.
Their experience and shared vision on the need to align technological development with European ethical and democratic values are fundamental.
Furthermore, the use of supercomputers in European infrastructures – such as Mare Nostrum in Spain and Leonardo in Italy – ensures that the training of these models is carried out under strict security and digital sovereignty standards.
Too little, too late?
For years, Europe has stood out for its ability to regulate emerging technologies. However, this trend has been criticized by some sectors who consider that excessive bureaucracy can limit innovation and delay the development of globally competitive solutions.
With OpenEuroLLM, the EU is committed to changing this paradigm, promoting an innovation ecosystem that allows the development of disruptive technologies without giving up the necessary safeguards to protect fundamental rights.
Will he make it? Considering the modest investment, the scope of the project does not appear to be on par with large private investments in Asia and North America.
Hopefully the project could demonstrate that it is possible to transform regulatory challenges into growth opportunities.
By promoting collaboration between universities, research centers and technology companies, OpenEuroLLM could generate a multiplier effect on innovation and position Europe as a “somebody” in AI.
Europe is not playing in the same league
Europe faces tough competition in the field of AI, especially from powers such as the United States and China.
The recent success of DeepSeek, an open language model developed in China, has highlighted the need for the continent to strengthen its investment in disruptive technologies.
In this context, investment in initiatives such as OpenEuroLLM is at least one step to try to regain ground in the global technology race.
However, the implementation of the AI Act means that companies and developers must operate in a complex regulatory environment.
The key for projects like OpenEuroLLM to thrive will be to find a balance between regulatory compliance and the ability to innovate and adapt to a constantly evolving global market.
This post is also available in: