The European Union has opened the bidding process for up to seven AI "gigafactories," a procurement program designed to mobilize more than €30 billion in public and private investment. The initiative aims to give European firms access to the computing power required to develop cutting-edge AI models, reducing the bloc's reliance on cloud services offered by U.S. companies such as Amazon Web Services, Microsoft Azure, and Google Cloud, which together account for a large share of Europe's cloud infrastructure market.

The decision comes as computing capacity has emerged as one of the most significant bottlenecks in the AI industry. Advanced AI models require clusters of tens of thousands of high-performance accelerators for training, and European officials acknowledge that existing public infrastructure is under severe strain. The move also comes amid intensifying global competition, with the United States and China both channeling substantial public and private resources into AI computing infrastructure.

The initiative forms part of the European Commission's broader AI Continent action plan, which seeks to establish sovereign AI computing capacity. It also reflects a global trend identified by Stanford University's 2026 AI Index, which found that countries worldwide are increasingly treating AI infrastructure as a strategic national asset on par with semiconductors, energy, and data.

The Hardware Imperative

A senior European Commission official summarized Europe's challenge plainly: "The current infrastructure is saturated in demand."

This means the primary obstacle to AI advancement in Europe is not a shortage of research talent but rather insufficient computing capacity. The assertion aligns with Stanford's conclusion that frontier AI development tends to favor entities with the financial resources to build and operate massive GPU farms.

Under the AI Gigafactories plan, each site will have the capacity to install up to 100,000 cutting-edge AI processors, supported by advanced networking, cloud software, reliable electricity sources, and state-of-the-art cooling systems for training and deploying large-scale models.

Europe has already established 19 AI Factories connected to public supercomputers. The gigafactories are intended to complement these by creating larger, private-led computing facilities. The goal is to ensure AI development occurs within Europe's established regulatory and data-governance frameworks, including the EU AI Act, the world's first comprehensive legal framework for artificial intelligence.

The initiative also aligns with the EU's Coordinated Plan on Artificial Intelligence, which since 2018 has sought to pool member-state investments rather than pursuing fragmented national efforts.

Funding Breakdown

The financing structure uses public money to attract private-sector investment. Of the total €30 billion, approximately €10 billion is expected from national and EU sources, leveraging an additional €20 billion from private investors. Public financing will generally be capped at 35% of project costs.

Some funding remains contingent on upcoming EU budget negotiations. The European High-Performance Computing Joint Undertaking (EuroHPC JU) will manage the competitive selection process, continuing the EU strategy of using public funding to catalyze private investment in critical technologies.

The European Commission press release and a separate announcement confirm the European Investment Bank Group's involvement in supporting the program.

Two-Lot Tender Structure

The tender is divided into two categories. One lot supports up to four medium-sized gigafactories, while the other covers as many as three large facilities. Depending on the project, combined EU and national government funding can range from €1 billion to €2 billion.

Medium-sized sites are expected to deploy 25,000 to 75,000 AI processors efficiently, while the largest facilities could exceed 100,000. Germany, Italy, Greece, Portugal, and Spain are among the countries expressing interest in hosting the major facilities.

A cluster of this scale would place an EU gigafactory among the largest AI training setups operated by major U.S. corporations, though it would still trail the massive GPU deployments planned by companies such as xAI. The comparison underscores Brussels' objective of building facilities capable of training frontier AI models rather than serving as conventional research supercomputers.

The Dependency Europe Cannot Eliminate

Building independent AI infrastructure does not, by itself, reduce Europe's dependence on foreign semiconductor manufacturers. Following the EU-U.S. trade agreement reached in July 2026, the Commission signed letters of intent with AMD, Nvidia, and Qualcomm to improve hardware access for the gigafactory projects.

Officials remain pragmatic about this dimension. Future hardware upgrades are expected to involve a broader range of suppliers as competition intensifies, but chips alone will not resolve the challenge. The International Energy Agency has emphasized that AI data centers are driving rising global electricity demand, making reliable power sources, sufficient grid capacity, and advanced cooling systems just as critical as the processors themselves.

Timeline

Applications close on 12 November 2026. EuroHPC evaluations will determine award decisions, expected in early 2027. Selected projects should begin operations approximately 18 months after contract signing.

Henna Virkkunen, Executive Vice-President of the European Commission for Tech Sovereignty, Security, and Democracy, called the launch "a milestone" for the bloc's AI ambitions.

The gigafactories are intended to do more than expand cloud capacity. EU officials envision the facilities providing infrastructure to accelerate AI adoption across manufacturing, healthcare, transportation, and public services. However, OECD analysis indicates that AI adoption across Europe remains uneven, meaning the €30 billion in additional computing power represents just one component of the broader effort to strengthen economic competitiveness.