A French consortium of technology, telecom, and infrastructure companies is seeking European funding for a large-scale AI data center project in France. The initiative, expected to cost around 10 billion euros, is intended to make a significant contribution to Europe’s ambition to reduce its dependence on American and Chinese AI infrastructure.
The consortium operates under the name AION and includes Capgemini, Orange, Iliad subsidiary Scaleway, investor Ardian, and energy group EDF. Reuters reports that the group intends to apply for funding from the new €20 billion European AI infrastructure fund announced by the European Commission late last year.
According to Benoît Gaillochet, head of infrastructure investments at Ardian, the French project alone could account for about half of the total European fund. The financing is ultimately expected to come from multiple sources, including private investors, bank loans, and European subsidies.
European AI Gigafactory
The project aligns with the growing European focus on digital sovereignty. Brussels is seeking to reduce dependence on American hyperscalers and Chinese cloud companies by building more of its own AI capacity. Within European policy circles, such large-scale infrastructure projects are increasingly being described as AI gigafactories.
This involves not only the development of AI models but, above all, access to sufficient computing power, energy, and data center infrastructure. European countries are trying to catch up in this area as AI training and inference require ever-increasing amounts of power and specialized hardware.
Scaleway CEO Damien Lucas outlines ambitious plans for the project. The ultimate goal is a data center environment with a total capacity of one gigawatt. This would roughly double France’s available data center capacity. According to him, the first construction phase would likely amount to around 100 megawatts.
Such capacities are needed to meet the growing demand for AI training and inference. European countries are trying to regain ground against hyperscalers and AI providers from the United States and China, which have invested billions in new AI data centers over the past few years.
EDF helps with power connections
A major bottleneck for new AI data centers remains the availability of energy and grid capacity. EDF is trying to accelerate that process by making former industrial sites that are already directly connected to the high-voltage grid available. This allows new data center locations to become operational more quickly.
The plans illustrate how AI infrastructure is increasingly becoming a key component of European industrial policy. Not only software and models, but also computing power and energy supply are now considered strategic assets in the international AI race.