Will an AI assistant understand you if you ask a question in Basque, Maltese, or Romani? Most Europeans do not speak English as a first language, yet modern AI systems are predominantly built on English-language data. Advancing open innovation to strengthen linguistic diversity and create value for Europe’s citizens and businesses is at the heart of the Microsoft Open Innovation Center’s (MOIC) mission.
Click to register and join policymakers, researchers, cultural institutions, and LINGUA grantees to examine how open innovation can strengthen Europe’s digital sovereignty
A gap worth closing
More than 60 regional and minority languages are spoken across the European Union, carrying centuries of cultural memory and local identity. Yet in the AI systems being built and deployed right now, most of these languages remain significantly underrepresented.
The reason is structural. AI models learn from data. Languages with smaller digital footprints (fewer digitised books, fewer websites, fewer hours of recorded audio) produce models that perform poorly, or not at all. The result is a self-perpetuating cycle: less data leads to weaker tools, weaker tools reduce digital activity in that language, and reduced activity means even less data to learn from.
The opportunity, however, is equally measurable. Only 20 percent of EU businesses currently use AI, a signal of how much room there is to grow. The European Commission estimates that breaking down language barriers in the Single Market could unlock up to €360 billion in intra-EU trade. AI that truly comprehends Europe's linguistic diversity, history, and values has the potential to serve EU citizens and power its economic future. For underrepresented communities, the difference is concrete: access to public services, education, healthcare, and economic opportunity in the language they speak.
Open by design
Solving a data challenge of this scale requires a collaborative effort, drawing on expertise distributed across the entire region. The only approach that effectively scales is an open one.
AI that truly comprehends Europe's linguistic diversity, history, and values has the potential to serve EU citizens and power its economic future
Through the LINGUA programme, MOIC and Microsoft’s AI for Good Lab partner with universities, research institutions, and language communities across Europe to build high-quality training datasets for low-resource languages. Launched in September 2025 as part of Microsoft's European Digital Commitments, a set of pledges to ensure that global technology serves Europeans and operates under European values and rules, LINGUA treats linguistic diversity as a strategic priority. Our first cohort spans 16 languages across 10 countries, including Maltese, Luxembourgish, Basque, Romani, and several regional dialects. This effort covers communities of more than 65 million speakers. The work of the first grantees ranges from speech recognition tools for endangered languages to AI-assisted archiving of oral traditions that might otherwise be lost within a generation.
Everything produced is open by design. The datasets, models, and tools are non-proprietary and available to any institution, developer, or competitor. Europe’s languages and cultures are not assets to be optimised: they are inheritances to be preserved, expanded, and shared. Open data means researchers and startups build on common foundations rather than duplicating fragmented efforts. Open models mean these systems can be freely adapted and extended. Open collaboration means public broadcasters, cultural institutions, universities, and technology companies working side by side rather than in separate silos. The goal is shared infrastructure, built for everyone.
Community-led innovation
Tools are necessary but not sufficient. They need the right conditions to reach the people who need them most. Among the LINGUA grantees, a team in Luxembourg is building an AI education platform in Luxembourgish, piloted directly in local schools. In Italy, researchers at the University of Naples are developing datasets to protect Neapolitan, Sicilian, and Roman dialect speakers from AI-generated health misinformation, in languages that mainstream models struggle to understand. Working alongside communities as active participants, letting their expertise and priorities shape the technology from the outset, is at the heart of MOIC's approach.
Working alongside communities as active participants, letting their expertise and priorities shape the technology from the outset, is at the heart of MOIC's approach
This partnership reflects a broader conviction: Europe’s ambition to lead on AI is most compelling when that leadership extends to all of Europe’s languages and communities. AI diffusion, business competitiveness, and the resilience of public services all depend on technology that works for citizens in the languages they actually speak. Linguistic diversity is not a cultural footnote to Europe’s digital agenda. It is one of its greatest strengths.
Where the work comes together
On 15 and 16 June in Strasbourg, MOIC and GitHub will bring together policymakers, researchers, cultural institutions, and LINGUA grantees for a two-day exchange at the Council of Europe, on one central question: how open innovation can actively strengthen linguistic and cultural representation?
Europe’s AI future will be shaped by choices made in the next few years. At MOIC, we are convinced that linguistic diversity is one of Europe's greatest competitive advantages, and open innovation is how we build on it. Strasbourg is where that conviction meets the work.
Sign up to The Parliament's weekly newsletter
Every Friday our editorial team goes behind the headlines to offer insight and analysis on the key stories driving the EU agenda. Subscribe for free here.