SoftBank says it will invest up to €75 billion to build French data centers

Background and Context

On May 30, 2026, SoftBank Group formally announced a landmark investment initiative, committing up to €75 billion to develop and operate a maximum of 5 gigawatts (GW) of additional data center capacity in France. This announcement marks a significant escalation in the global competition for artificial intelligence infrastructure, positioning SoftBank as a major player in the European market. The scale of this investment is unprecedented for a single-country infrastructure project by SoftBank and ranks among the largest in French history. The primary objective of this massive capital deployment is to address the exponential growth in computing power required for AI training, inference, and cloud computing services across Europe. With a capacity of 5 GW, the project is designed to support hundreds of thousands of high-performance AI clusters, fundamentally altering the existing supply structure of computational resources in the region.

The strategic rationale behind this move is rooted in the consensus that "compute is power." As large language models continue to break through parameter size barriers, traditional data center architectures are no longer sufficient to meet the demands for low latency and high throughput. SoftBank’s approach involves deploying clusters based on the latest GPU architectures, which require extremely high power density and cooling efficiency. This necessitates a transformation of data centers from standard IT facilities into energy-intensive industrial complexes. The company has outlined a phased implementation plan, prioritizing locations with stable energy supplies and robust network connectivity potential. This strategy aims to build a highly scalable, low-carbon foundation for AI computing, reflecting SoftBank’s broader strategic response to the global AI infrastructure wave through its Vision Fund vehicles.

Deep Analysis

From a technical and business model perspective, SoftBank’s investment represents a shift from pure software investment to heavy asset ownership and energy binding. The company is adopting an integrated "build-and-operate" model, which not only generates stable cash flow through the sale or leasing of server racks but also involves deep engagement in upstream power procurement and downstream compute distribution. This model resembles that of telecommunications operators, but serves the "compute" of the digital age. By providing infrastructure that complies with European standards, SoftBank is building a moat against the potential monopoly of American cloud giants in the European market. This move also provides European AI startups with a reliable source of compute, positioning SoftBank as a key ecosystem builder rather than just an investor.

The technical requirements of 5 GW of capacity pose significant challenges and opportunities. The deployment of such a vast amount of computing power requires not only advanced cooling systems but also a secure and abundant energy supply. SoftBank’s decision to focus on France is partly driven by the country’s strong nuclear energy infrastructure, which offers a stable and low-carbon power source suitable for AI workloads. However, the sheer volume of power required will likely necessitate new grid connections and potentially new energy generation capacities. The company must also navigate the complexities of data sovereignty laws, such as the General Data Protection Regulation (GDPR), ensuring that its infrastructure meets the strict privacy and localization requirements of European clients. This compliance-focused approach adds a layer of complexity to the operational model but also serves as a competitive advantage in a market increasingly concerned with data privacy.

Industry Impact

SoftBank’s entry into the French data center market is expected to have profound implications for the European technology sector. First, it intensifies the competition for critical resources such as land, electricity, and water. Data centers are known as the "water, electricity, and coal" of the digital age, and the 5 GW capacity will place significant stress on local grid stability. This pressure may force the French government to accelerate the construction of green energy infrastructure or reconsider the role of nuclear power in supporting AI computing. Second, the announcement is likely to trigger a chain reaction, attracting other tech giants such as Microsoft, Amazon AWS, Google Cloud, and European players like OVHcloud to increase their investments in France. This clustering effect could create a robust AI infrastructure ecosystem in the region.

For end-users, including European enterprises and research institutions, the increased supply of AI compute will provide more diverse and cost-effective options, reducing dependency on single suppliers. However, this could also lead to a structural rise in compute costs, as the high capital expenditure of building and operating these facilities is eventually passed on to consumers. Furthermore, SoftBank’s presence may alter the market share distribution in the European cloud market, potentially weakening the dominance of US-based cloud providers and promoting digital sovereignty. For SoftBank, this is a high-risk, high-reward strategic bet. Its success will depend on its ability to manage energy costs, navigate technological iteration risks, and mitigate uncertainties arising from geopolitical factors.

Outlook

Looking ahead, SoftBank’s data center construction in France will serve as a key indicator of the evolution of global AI infrastructure. A critical area of focus will be the sustainability of its power sources. In the context of carbon neutrality, ensuring a green power supply for 5 GW of capacity will determine the long-term compliance and social acceptance of the project. Additionally, the evolution of technology routes will impact investment returns. As AI model architectures are optimized and specialized chips become more prevalent, the energy consumption per unit of compute may decrease, but it could also increase due to the rising complexity of models. SoftBank must remain flexible in its hardware configuration strategy to adapt to these changes.

Geopolitical factors also cannot be ignored. Coordination between the US and Europe on digital trade rules may affect data flow and infrastructure investment. The implementation progress of the project will test SoftBank’s execution capabilities and financing strength. If the project is successfully completed, SoftBank is poised to establish a leadership position in the European AI infrastructure sector, providing a key pivot for its global AI strategy. Conversely, if it encounters policy resistance or technical bottlenecks, the substantial sunk costs could pressure SoftBank’s financial health. Ultimately, this investment is not just a commercial decision for SoftBank but a crucial step in the restructuring of the global AI compute landscape, warranting close attention from technology, energy, and policy makers.