
Major investment to strengthen France’s AI infrastructure, support European technological sovereignty and anchor advanced data center manufacturing in Dunkirk with Schneider Electric PARIS, June 1, 2026 — SoftBank Group Corp. has announced its commitment to develop and operate 5 GW of AI data center capacity in France, representing an investment of up to €75 billion. The […] The post SoftBank Group to Build 5 GW of AI Data Center Capacity in France appeared first on HPCwire .
The global race for AI leadership is intensifying, driving nations and major corporations to invest massively in the underlying compute infrastructure. SoftBank, with its focus on AI, strategically positions itself in key regional markets.
This investment signifies a major commitment to building dedicated AI compute capacity outside traditional hubs, bolstering European technological independence and setting a precedent for similar national-scale AI infrastructure projects. It highlights AI's dependence on scalable, dedicated power infrastructure.
Europe's AI compute capacity is significantly boosted, potentially reducing its reliance on US-based AI cloud providers, and demonstrating the critical link between AI development and the energy sector.
- · France (AI infrastructure, economy)
- · SoftBank Group
- · Schneider Electric
- · European AI developers
- · US-centric AI cloud providers (potential market share erosion in Europe)
- · Small-scale AI data center operators (increasing competition)
France becomes a major hub for AI data center development and operation in Europe.
This investment could spur further national and regional initiatives across Europe to build out sovereign AI compute capabilities and attract AI talent.
Increased energy demand from these data centers could accelerate investment in advanced energy solutions, potentially including small modular reactors or enhanced renewable grid infrastructure.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at HPCwire