Search

EU Plans Seven AI Gigafactories With €30 Billion as Von der Leyen Pushes Tech Sovereignty

By: IDCNOVARegion: Europe
European leaders used the bloc's first-ever AI summit on November 13, 2026, to signal an aggressive push toward computing self-sufficiency, with European Commission President Ursula von der Leyen announcing that seven AI Gigafactories are in planning and that roughly €30 billion is being mobilised to back them. The gathering, an informal European Council meeting, underscored how urgently Brussels wants to build homegrown computing capacity and reduce its reliance on foreign hyperscalers.

The stakes extend well beyond the summit chamber. Compute capacity determines which companies can train frontier models, how cheaply they can run inference, and where sensitive data ultimately resides. For businesses building or buying AI in Europe, the announcements carry direct implications for cost structures, latency, and data residency guarantees.

Von der Leyen said the planned gigafactories "will provide the computing capacities our startups need to build the most advanced AI models." She also laid out the scale of the gap Europe faces, stating that the bloc needs "at least €500 billion according to our own estimates," identifying chips, compute, and frontier AI as the critical missing pieces. She promised "an ambitious proposal" on how to finance that ambition. European Council President António Costa framed AI as "the most important technology of our time," and discussions centred on competitiveness and reducing dependence on outside players.

The rationale is straightforward: Europe generates more tech start-ups than the United States, yet scaling large models demands massive GPU clusters that today sit overwhelmingly with US hyperscalers. The gigafactories are intended to close that gap, giving startups from Lisbon to Warsaw the ability to train serious models on European soil without competing for scarce capacity abroad. But €30 billion across seven sites represents a starting point rather than a finish line, and the €500 billion estimate illustrates how far the ambition stretches beyond currently committed funding.

Adoption data presented at the summit complicates the conventional narrative of Europe as a regulator rather than a builder. Von der Leyen said 37% of EU businesses used AI on average in 2025, with Finland reaching 66%, which she described as almost double the US rate. Those figures suggest European companies are adopting AI widely even as they lack domestic infrastructure and frontier models. The Commission is also focusing on five sectors where AI can add the most value, with the first phase of that work due in about two weeks, supported by Jim Snabe, appointed earlier this year as Special Envoy for Industrial Artificial Intelligence.

Governance remained a central theme. Costa argued that infrastructure alone would not suffice, stating that "without trust, AI strategy cannot succeed," and noted that the EU was the first region with an AI governance framework and should push standards through the G7, G20, and the UN. He also announced that AI and its social impact will be the central theme of the 2027 Social Summit. A day earlier, on November 12, the European Parliament adopted a resolution by 405 votes to 133, with 60 abstentions, calling for human-centred AI in EU institutions, prioritising European AI tools, and creating a budget-neutral "European Parliament AI observatory." Rapporteur Manolis Kefalogiannis said that "AI does not progress linearly. It advances in leaps, and Europe must take the lead in shaping it."

The financing question will ultimately determine the outcome. Public budgets alone cannot reach €500 billion, and von der Leyen pointed to completing the Savings and Investment Union, attracting private investment, and reflecting the ambition in the next long-term EU budget, while naming the Scaleup Europe fund as a success story worth building on. Costa called the next budget, particularly the European Competitiveness Fund, "decisive," and mentioned the proposed "28th regime," a single set of rules that would let start-ups operate across all member states without navigating 27 separate legal systems. For founders, that regulatory simplification may matter as much as access to GPUs, given how fragmented company law, tax, and hiring rules have historically slowed cross-border scaling in Europe. Affordable clean energy also looms as a quiet constraint, with Costa listing it as a condition for success and any gigafactory certain to draw substantial power, making cheap green electricity a likely factor in site selection.

For non-EU technology firms serving European clients, the implications are concrete. Sovereignty pushes typically reshape procurement first, and clients may begin asking where models run, which cloud regions are used, and whether deployments can sit on European infrastructure. Firms that design AI systems to be cloud-portable, with clear data flows and the ability to swap model providers, will adapt most easily, while those locked into a single provider and region will feel mounting pressure. Portability is becoming a sales advantage rather than a technical nicety.