SoftBank earmarks $88 bn for a 5‑GW AI data hub in France, signaling a strategic shift beyond the U.S.

SoftBank’s $88 bn AI data hub plan

SoftBank Group announced a commitment of up to 14 trillion yen (≈ $88 billion) to construct a new data‑centre campus in France. The project will deliver 5 gigawatts (GW) of power capacity, placing it among the largest AI‑focused facilities in Europe. The investment will be rolled out in phases, with the first phase targeting 1.5 GW of compute power by the end of 2027 and the full build‑out expected by 2032.

Market context: why Europe now?

1. AI compute scarcity in the U.S. – Recent spikes in demand for GPU‑heavy workloads have driven up spot prices on major cloud platforms by 30‑45 % year‑over‑year. Companies are looking for alternative locations to avoid bottlenecks and geopolitical risk.
2. European energy policy – France’s nuclear fleet provides low‑carbon baseload electricity, while the government has pledged €30 bn in subsidies for green data‑centre projects. The 5 GW target aligns with the EU’s goal of 20 GW of AI‑specific compute by 2030.
3. Competitive pressure – Google, Microsoft and Amazon have already secured large leases for French sites, and the French Tech Ministry is actively courting foreign investors to cement Paris as a continental AI hub.
4. SoftBank’s portfolio – The group’s recent earnings showed a 400 % increase in AI‑related revenue, driven by stakes in OpenAI, Anthropic and its own in‑house AI platform. Securing dedicated compute is now a prerequisite for scaling those services.

Strategic implications for SoftBank

• Diversification of compute assets – By locating a massive facility in Europe, SoftBank reduces reliance on U.S. data‑centre providers, mitigating supply‑chain and regulatory risks.
• Revenue upside – SoftBank plans to lease the French capacity to both internal AI ventures and third‑party cloud operators. At an estimated average lease rate of $9,000 per kW per year, the 5 GW campus could generate $45 bn in annual recurring revenue once fully occupied.
• Energy cost advantage – France’s regulated electricity price of €0.07/kWh is roughly 30 % lower than the German market, improving operating margins for AI workloads that consume over 200 MWh daily at peak.
• Regulatory goodwill – Partnering with the French government on a green‑energy‑linked project positions SoftBank favorably for future EU data‑sovereignty incentives, potentially unlocking additional tax credits worth up to €200 m per year.

What it means for the broader AI ecosystem

• European AI sovereignty – The hub provides a locally‑hosted alternative for European firms that face data‑privacy restrictions under GDPR. Companies can run large language models without exporting data to the U.S., a factor that could accelerate AI adoption in regulated sectors such as finance and healthcare.
• Pressure on U.S. cloud pricing – As capacity in Europe becomes more abundant, U.S. providers may be forced to adjust pricing or accelerate the rollout of new GPU‑dense regions to retain market share.
• Energy‑intensive AI scrutiny – With 5 GW of power demand, the project will be a test case for integrating renewable offsets and demand‑response mechanisms. Success could set a benchmark for sustainable AI compute at scale.
• Competitive response – Expect rivals such as Alibaba Cloud and Tencent to announce similar European expansions, potentially sparking a wave of cross‑border data‑centre investments.

Bottom line

SoftBank’s $88 bn French AI data hub is more than a capital outlay; it is a strategic hedge against U.S. compute bottlenecks, a lever for monetising its AI portfolio, and a catalyst for Europe’s push toward AI sovereignty. If the project meets its 5 GW target on schedule, SoftBank could secure a $45 bn annual lease pipeline, while reinforcing its position as a global AI infrastructure provider.