Europe initiates bidding for seven AI 'gigafactories' in a €30 billion effort to keep pace.

      The European Commission has initiated the bidding process for the establishment of up to seven AI “gigafactories,” which are extensive computing centers designed to empower the bloc to train leading AI models independently from the United States.

      The tender call, announced today, represents an approximate €30 billion initiative aimed at bridging the gap with AI advancements in the US and China. The funding structure is divided between public and private investments. Brussels and its member states are anticipated to contribute about €10 billion collectively, while the remaining €20 billion is expected to come from private investors, with the hope that public funds will attract the necessary capital for these initiatives.

      The scale of the project is intentionally vast. Each gigafactory is expected to contain at least 100,000 state-of-the-art AI chips, making them roughly four times more powerful than the largest data centers currently operating in the EU, thereby more than doubling the bloc’s AI computing capacity.

      The term “gigafactory” marks a significant upgrade. Europe is already financing a network of 19 smaller “AI factories” linked to its supercomputers; however, the new facilities are designed to be significantly larger, built to train the advanced models that only a limited number of US and Chinese labs can currently support.

      There is clearly a strong interest in this initiative. A prior call for expressions of interest received 76 responses from various consortia throughout Europe, with ten countries, including Germany, France, Italy, Spain, and Poland, eager to host, and France already indicating it may pursue this independently.

      The motivation behind this move stems from competitive concerns. Europe has observed the US and China heavily invest in AI infrastructure while EU companies rely on computing power from American cloud services. Henna Virkkunen, the Commission’s tech-sovereignty chief, referred to the substantial computing power as “a strategic necessity.”

      Nonetheless, the ambitious plans precede the actual funding. As of now, only about €1 billion of Brussels’ contribution is confirmed; the remainder hinges on the EU’s forthcoming long-term budget, known as the Multiannual Financial Framework, which has yet to be finalized.

      A senior official candidly noted that decisions regarding the next MFF cannot be anticipated, describing the amount as a “best estimate” rather than assured funding.

      The timeline appears tight given the uncertainties. Construction is slated to commence in early 2027, with the first gigafactories projected to be operational by mid-2028, which relies on budgets, site acquisitions, and supply chains aligning.

      Historically, caution is advised. The concept of gigafactories was initially proposed in early 2025 and has since experienced multiple delays, frustrating some key partners necessary for the construction.

      There is also a paradox at the core of the sovereignty argument. The chips for these European gigafactories are predominantly sourced from American companies like Nvidia, AMD, and Qualcomm, implying that the bloc would effectively be purchasing its independence from the very firms it seeks to distance itself from.

      Energy costs are another limiting factor. Electricity in Europe is two to three times more expensive than in the US or China, and the power requirements for data centers are already a significant constraint that cannot be swiftly alleviated by subsidies.

      Critics have challenged the entire model, suggesting that renting sovereign-branded capacity built on foreign-owned hardware creates an illusion of independence rather than true autonomy, asserting that Europe’s primary deficiency lies in chip production and model development, not physical infrastructure.

      In response, proponents argue that infrastructure is an essential first step. While computing power alone won't create a European equivalent of OpenAI, the absence of it leaves startups without the capacity to scale their training efforts, which is what the Commission is attempting to address, albeit at a high cost.

      The Commission maintains that action must be taken at some point. Public funding entities would receive a proportional share of compute resources for research and public initiatives, with the expectation that building this infrastructure will attract talent, startups, and eventually chip manufacturing to Europe.

      Currently, the bidding process is open, although the funding largely remains hypothetical. Firms are invited to compete to shape Europe’s AI future, with an understanding that the continent has expressed ambitions of €30 billion while securing only about a thirtieth of that amount thus far.