Europe’s Open-Source AI Landscape: A lever for Innovation and Sovereignty

A new report on the European Open-Source AI Landscape reveals where the EU stands today and how scaling adoption of Open Source AI can boost competitiveness and strengthen digital sovereignty.

Open-source AI includes models, tools and datasets whose components — like source code, model weights and documentation — are openly available to use and modify. This openness lowers barriers for universities, public institutions and business – big and small - to deploy AI without relying on proprietary AI models and systems.

The report shows that open-source is already widely used across the AI tech stack: over half of developers regularly rely on open models, datasets and tools. Its transparency also supports safer AI by providing valuable information to researchers to study the safety of these models in detail.

A fast-moving global race

Since 2022, the number of publicly released models has more than doubled, and more of them now come with open weights. This gives developers far more transparency and control. Moreover, such open models are quickly closing the gap with proprietary systems and, on some benchmarks, even matching expert-level performance.

Europe’s strengths: collaboration and trust

Europe has strong assets: top research, close academia–industry collaboration, and major contributions to tools like scikit-learn, spaCy and PyTorch that underpin thousands of commercial AI systems globally. In addition, many European AI startups are releasing open-weight models under open-source licences that others can build on.

Adoption of AI in European companies, however, remains low, with 14% of EU firms using AI in 2024. In response to this the EU’s Apply AI strategy, launched in October 2025, is focused on stimulating greater uptake of AI by EU businesses.

According to the report, open-source AI can help close this ‘use gap’ by offering transparent, reusable and cost-effective tools that make it easier for organisations to deploy trustworthy AI aligned with EU values.

Europe’s Strategic Opportunity in Open-Source AI

Nevertheless, compute access remains a barrier for European innovators. The 19 EU-funded AI Factories and expanded EuroHPC supercomputers aim to address this by giving startups and SMEs free access to GPU capacities needed to develop European models, including open-weight models. This ‘public good’ infrastructure for the AI age is already delivering - in September 2025 the  Latvian SME Tilde launched TildeOen LLM, a 30 billion parameter open-source language model trained using 2 million GPU hours on the EuroHPC LUMI supercomputer.

Europe’s Path to a Sovereign and Competitive AI Future

Thanks to open-source, AI is rapidly becoming a low-cost, widely accessible technology. This, coupled with inference costs dropping by more than 99% in two years, creates an opportunity for Europe to strengthen its position. Rather than competing exclusively to build the largest frontier models, the report highlights that Europe can lead by developing open, trusted, multilingual and sector-specific AI that aligns with its industrial priorities and societal values. Achieving this will require increased investment, improved access to compute, and targeted support to help organisations, particularly SMEs, adopt and deploy open-source AI effectively.

Read the full report!