The European Union is moving decisively to strengthen its place in the global artificial intelligence race. On 30 July 2026, the European Commission announced a call for tenders to establish up to seven AI Gigafactories across Europe, presenting the initiative as a major step toward technological sovereignty and its wider ambition to become an “AI Continent.” The plan is designed not only to expand raw computing power, but also to give Europe more control over how advanced AI systems are built, trained, and deployed.
This new phase builds on the existing AI Factories network and reflects a broader policy shift: sovereign AI models are no longer a side objective, but a formal strategic goal. From frontier model development to sector-specific systems for energy and industry, Brussels is linking compute infrastructure, cloud capacity, and industrial policy into one coordinated push. In that context, the selection of AI factories and gigafactories is about far more than data centres; it is about creating the foundation for European AI independence.
A New Push for European AI Capacity
The Commission’s July 2026 announcement framed AI Gigafactories as a direct answer to Europe’s need for stronger AI infrastructure. According to the official press release, the initiative is intended to “expand Europe’s AI computing capacity” and to boost the region’s ability to support advanced model training and deployment. That framing matters, because access to large-scale compute has become one of the main bottlenecks in developing leading AI systems.
The financial ambition is equally striking. The program is expected to be backed by up to €10 billion in combined EU and national public funding, while also unlocking at least €20 billion in private investment. This blend of public support and market participation shows that the EU wants to de-risk strategic infrastructure while also encouraging a wider commercial ecosystem around sovereign AI development.
Politically, the message is clear: Europe does not want to remain dependent on foreign platforms for the next generation of artificial intelligence. The Gigafactories initiative is being presented as part of a wider strategic effort to reinforce resilience, strategic autonomy, and long-term competitiveness. In other words, compute is now being treated as a pillar of sovereignty.
What AI Gigafactories Actually Are
According to the European Commission, AI Gigafactories are large-scale facilities designed for the development and training of next-generation AI models with trillions of parameters. These are not ordinary computing centres. They are envisioned as highly specialized environments able to handle the extreme demands of frontier AI, including huge processing loads, rapid data movement, and the energy requirements associated with training very large models.
Each gigafactory is expected to combine more than 100,000 advanced AI processors with strong power supply, high-performance networking, energy-efficiency systems, and automation capabilities. That technical specification reveals the scale of the EU’s ambition. It also underlines why this initiative is being compared to industrial infrastructure rather than standard digital investment.
The facilities are meant to support the full lifecycle of advanced AI systems. The Commission says they will provide infrastructure for training, inference, and fine-tuning frontier AI models, opening access to startups, scale-ups, SMEs, industry, academia, and public authorities. This broad access model is important because it suggests the EU wants these resources to serve an entire innovation base, not just a few dominant players.
How AI Factories Became the Foundation
The Gigafactories plan did not appear in isolation. It rests on the earlier buildout of the EU’s AI Factories network, which now counts 19 AI Factories according to the Commission’s dedicated page. These facilities form the base layer of Europe’s AI compute strategy, giving researchers and companies access to AI-optimised supercomputing resources and helping create the practical experience needed for larger sovereign AI programs.
The network was expanded in stages. The Commission announced additional AI Factories in March 2025 and then followed with another wave in October 2025, extending the footprint across several member states. That staged rollout suggests a deliberate approach: first create a distributed network of accessible AI infrastructure, then scale toward much larger gigafactory capacity for frontier systems.
Funding has already been substantial. The Commission says the EU and participating EuroHPC countries have committed more than €2.6 billion to the AI Factories and Antennas initiative. That investment has not only supported core computing resources, but also the ecosystem around them, helping the Union transition from scattered national efforts toward a more coordinated European capability.
Regional Access and the Role of Antennas
One challenge in any large-scale European infrastructure plan is avoiding concentration in only a handful of countries or institutions. To address that, the Commission has also promoted AI Factory Antennas, which act as regional access points. There are now 13 of these antennas, and their role is to connect national and local AI communities to the broader European supercomputing network.
These antennas matter because sovereign AI cannot be built only in capital cities or major technology hubs. If the EU wants startups, universities, SMEs, and public agencies across the bloc to participate, access has to be geographically and institutionally broadened. The antenna model helps make advanced AI infrastructure more reachable for users who may not operate near the main computing centres.
This regional dimension also supports policy cohesion. By widening participation, the EU increases the chance that sovereign models will reflect Europe’s diverse languages, regulatory approaches, industrial needs, and public-interest priorities. In that sense, the antenna network is not just an administrative add-on; it is part of the social and economic architecture of digital sovereignty.
Sovereign Models as a Formal Policy Goal
The development of sovereign AI models is now deeply embedded in official EU strategy. The Horizon Europe 2026-2027 work programme states that European support will help develop sovereign frontier AI with safety by design, including multimodal and agentic capabilities. That language indicates the Union is not merely trying to adopt existing AI tools, but to participate in the creation of the most advanced systems.
This ambition is reinforced by wider strategy documents that commit the Commission to supporting open frontier AI models and sovereign frontier AI capabilities. The emphasis on openness is notable. Europe appears to be positioning itself as a region that can combine competitiveness with public-interest principles, using infrastructure investment to support alternatives to closed, externally controlled AI ecosystems.
The significance of this policy shift should not be understated. When sovereign frontier models become a formal objective, infrastructure choices such as AI factories and gigafactories gain a new strategic role. They are no longer simply research assets; they become instruments for shaping who owns, governs, and benefits from advanced AI in Europe.
Why Technological Sovereignty Is Driving the Strategy
The Commission has explicitly framed the AI factories agenda as part of Europe’s push for technological sovereignty. In its statements, Brussels has argued that the new infrastructure will strengthen resilience and strategic autonomy across the Union. This reflects a wider concern that dependence on non-European compute, cloud platforms, chips, or foundational models could leave the bloc exposed economically and politically.
That thinking was further integrated into the European Technological Sovereignty Package proposed on 3 June 2026. Covering semiconductors, AI, cloud, and open source, the package linked these areas under the broader objective of digital autonomy and resilience. AI factories therefore sit within a larger ecosystem policy, one that sees compute, software, and industrial capability as interdependent.
The same logic appears in the proposed Cloud and AI Development Act, or CADA. The Commission has argued that Europe needs more cloud and data-centre capacity to support the diffusion of AI alongside AI factories and gigafactories. Put simply, sovereign AI models require more than training clusters: they need an entire stack of infrastructure that remains available, scalable, and governed within Europe.
From Frontier Research to Industrial and Public Uses
One reason the AI factories strategy has gained momentum is that the Commission is tying it to practical use cases. In the energy sector, for example, the EU has said it should build sovereign, secure AI models trained on European data and developed by EU firms to support grid management and planning. This provides a concrete illustration of how sovereign AI can intersect with critical infrastructure policy.
That vision is already taking shape through AI.grids, launched in June 2026 with 48 partners. The project aims to develop EU sovereign AI models for improving grid management and planning, showing how advanced infrastructure could be used to support highly specialized public-interest systems. It also demonstrates that sovereign models are not only about prestige or competition with foreign labs, but about trusted applications in essential sectors.
Beyond energy, the same logic can extend to healthcare, manufacturing, science, administration, and defense-adjacent resilience planning. By giving startups, industry, researchers, and public authorities access to shared compute, the EU is trying to create conditions where sovereign models can emerge across many domains. This could help Europe move from being primarily a regulator of AI to becoming a stronger builder of AI systems as well.
What Comes Next for the AI Continent Ambition
The EU’s AI Continent Action Plan says the bloc aims for at least 13 operational AI factories by 2026 and plans to triple data-centre capacity over the next 5.7 years. Those targets show that AI factories are only one layer of a much larger infrastructure transformation. Europe is attempting to align compute expansion, industrial strategy, and sectoral AI adoption within a compressed timeline.
The selection and rollout of gigafactories will therefore be closely watched. If the program succeeds, Europe could significantly improve its ability to train and deploy sovereign AI models at scale. If implementation is slow, fragmented, or constrained by power, supply-chain, or permitting issues, the gap with global AI leaders may remain difficult to close.
Still, the direction of travel is unmistakable. By choosing AI factories and launching gigafactories, the EU is sending a signal that access to compute is now central to economic power, public resilience, and digital independence. The real test will be whether this infrastructure can translate policy ambition into world-class AI systems built in Europe and for Europe.
In that sense, the AI factories strategy marks an important evolution in the European approach to technology. For years, the EU was often characterized mainly as a rule-maker, shaping standards for privacy, competition, and platform governance. Now it is also trying to shape the underlying industrial base of AI, using investment and coordinated infrastructure to support sovereign models and broader strategic autonomy.
The success of this effort will depend on execution, inclusion, and sustained financing. But the objective is already clear: to ensure that Europe has the computing capacity, institutional support, and policy coherence needed to develop its own advanced AI future. As the Union selects and expands AI factories, it is laying the groundwork for a more sovereign place in the global AI landscape.
