SoftBank Ignites Europe’s AI Future: €75BN French Hub Becomes Continent’s Largest

Stay informed with free updates

Simply sign up to the Artificial intelligence myFT Digest — delivered directly to your inbox.

SoftBank, under the visionary leadership of Masayoshi Son, has made an monumental pledge to invest up to €75bn in establishing a vast network of AI computing clusters across France. This commitment signals a profound strategic pivot for the Japanese conglomerate, moving aggressively into foundational AI infrastructure, and is poised to underpin what would become Europe’s largest data centre project. The investment reflects a growing consensus that robust, large-scale compute power is the bottleneck for AI development, and positions Europe in a crucial race to close the significant infrastructure gap with global leaders like the US and China.

This unprecedented commitment represents the largest single AI infrastructure investment by Masayoshi Son’s SoftBank group outside of North America, underscoring a global strategy to dominate the backbone of the AI revolution. For French President Emmanuel Macron, the announcement is a significant geopolitical and economic victory, providing a substantial boost ahead of his annual “Choose France” summit. This high-profile event, designed to attract foreign direct investment and showcase France’s business-friendly environment, now boasts a marquee deal that validates Macron’s efforts to position the nation as a premier destination for advanced technology.

The swift materialisation of this deal, following an intimate dinner between President Macron and Mr. Son in Tokyo in early April, highlights the high stakes and rapid pace of global AI competition. Sources familiar with the discussions indicate that Macron’s compelling pitch centred on France’s unique combination of abundant, low-carbon nuclear power and a streamlined regulatory pathway for AI facility development. This pragmatic approach addresses two critical pain points for hyperscale data centre projects: sustainable and affordable energy, and the often-protracted planning and permitting processes that can delay deployment by years in other jurisdictions.

Masayoshi Son, SoftBank’s founder and chief executive, articulated the strategic rationale, stating, “SoftBank is proud to make this major commitment to France. With its industrial capabilities, talent base and national ambition, France is uniquely positioned to become a leading AI infrastructure hub in Europe.” This statement echoes SoftBank’s assessment of France’s long-term potential, not just as a power source, but as a holistic ecosystem capable of supporting the immense operational and human capital requirements of next-generation AI compute, potentially shifting the centre of gravity for European AI development.

Despite this significant win, the broader European landscape continues to lag considerably behind the US, China, and even emerging hubs in the Middle East when it comes to the development of hyperscale data centres essential for meeting the exponentially growing demand for AI computing power. Historically, investment flows have been diverted to regions offering more attractive energy costs, expedited grid connections, and less onerous regulatory environments concerning planning, data sovereignty, and carbon emissions. This SoftBank deal, therefore, represents a crucial litmus test for whether Europe can effectively counter these systemic disadvantages and carve out a competitive niche in the global AI infrastructure arms race.

SoftBank’s phased commitment begins with an initial investment of €45bn, earmarked for the development of 3.1 gigawatts (GW) of compute capacity in the northern Hauts-de-France region by 2031. This ambitious first stage is slated to be followed by an additional 2GW of expansion, bringing the total planned capacity to an astounding 5GW. This multi-phase approach allows for flexible capital deployment and adaptation to evolving market demands, while still signalling a long-term strategic vision for sustained growth.

A cornerstone of this initiative will be a major facility in Dunkirk, where SoftBank plans to collaborate with industrial giant Schneider Electric. This partnership aims to establish not just a data centre but a comprehensive hub for AI infrastructure and advanced robotics manufacturing. The choice of Dunkirk is strategically astute, leveraging its geographic proximity to major European economic centres like London, Brussels, and Amsterdam. This prime logistical location minimises latency and maximises connectivity, making it an attractive proposition for enterprise and cloud customers seeking high-performance, low-delay AI services across the densely populated North-Western European corridor.

Should the full 5GW complex be realised, its scale would be truly transformative. To put this into perspective, 5GW is roughly equivalent to the continuous output of five large nuclear power stations, or the peak electricity demand of an entire metropolis like New York City. This staggering energy requirement underscores the profound shift in global power consumption driven by AI, and SoftBank’s willingness to commit approximately €75bn (or $87bn at current exchange rates) to this endeavour demonstrates a conviction in the long-term, high-return potential of owning this critical utility of the digital age. It also highlights the tremendous pressure this will place on local and national energy grids, even in a nuclear-powered country like France.

The French commitment is not an isolated venture but a critical component of SoftBank’s rapidly expanding global AI infrastructure portfolio. This includes a previously announced, even larger, 10GW data centre project in Ohio, unveiled by Trump administration officials in March, which further illustrates the strategic importance SoftBank places on establishing compute capacity across key economic blocs. These parallel mega-projects signal a coherent, multi-continental strategy to build a resilient and geographically diverse network of AI processing power, hedging against regional risks and catering to varied regulatory landscapes.

Indeed, SoftBank’s overarching corporate strategy has coalesced around establishing an indispensable position at the heart of the AI revolution. Beyond infrastructure, the group has reportedly committed over $60bn of investment into OpenAI, the progenitor of ChatGPT, demonstrating a full-spectrum approach from application layer to foundational compute. This is further complemented by plans for public listings of its robotics and energy businesses in the US, and a concerted effort to build up semiconductor capacity around its crown jewel, the UK chip designer Arm. This integrated ecosystem approach aims to capture value across the entire AI value chain, from hardware and power to core algorithms and applications.

Son’s escalating ambitions for independent, globally distributed data centres appear to have, to some extent, superseded or at least diversified from earlier plans for the colossal $500bn “Stargate” joint venture. Stargate was initially conceived to provide compute at an unprecedented scale, primarily for OpenAI’s exclusive use. The shift toward a more diversified, multi-customer infrastructure model suggests a broader market play, positioning SoftBank as a foundational provider for numerous AI developers and enterprises rather than a single dominant partner. This reduces specific client dependency while potentially increasing overall market reach and revenue streams.

Industry estimates underscore the immense capital intensity of these projects, suggesting that every 1GW of AI infrastructure can cost roughly $50bn in total, encompassing land acquisition, construction, power generation and transmission, and the highly specialised IT equipment. This implies that SoftBank’s €75bn plan for 5GW will necessitate bringing in significant additional financing, likely from a consortium of as-yet-unnamed partners. This highlights the collaborative, capital-heavy nature of building next-generation AI foundations, often requiring alliances between technology firms, infrastructure funds, and sovereign wealth funds.

Understanding the financial architecture of such mega-projects is crucial for investors. Typical funding arrangements for infrastructure of this magnitude would see SoftBank providing a relatively modest initial equity financing injection. The overwhelming majority of the capital required would then be secured through sophisticated debt-based project financing, leveraging long-term contracts and asset-backed securities. This model allows SoftBank to deploy its equity strategically across multiple ventures while de-risking individual projects through diversified funding sources, albeit exposing partners to significant long-term debt obligations.

Further solidifying its global footprint, SoftBank is also an integral part of a powerful consortium planning another 5GW of AI infrastructure in Abu Dhabi. This Middle Eastern venture involves prominent players such as G42, OpenAI, Oracle, Nvidia, and Cisco, assembling a formidable alliance of technology, cloud, and chip giants. The breadth of these partnerships across three continents – North America, Europe, and the Middle East – paints a clear picture of SoftBank’s strategic intent to establish a ubiquitous, interconnected network of AI compute capacity, positioning itself as a central nervous system for the burgeoning global AI economy.

While the infrastructure itself is ambitious, the ultimate customers for the compute power and the specific providers of the advanced computing equipment (such as GPUs from Nvidia or AMD) for the French facility remain undisclosed. This opens up significant market opportunities for hardware manufacturers, cloud service providers, and AI developers. The eventual selection of these partners will be a key indicator of market dynamics and competitive positioning within the high-performance computing sector, influencing future supply chains and strategic alliances.

It is imperative to note, however, that the landscape of pledged AI investments is not without its cautionary tales. Even as technology titans commit hundreds of billions of dollars to expanding AI capacity, a notable proportion of projects, often trumpeted by politicians and executives with great fanfare, have failed to materialise or have been significantly delayed. This introduces an element of execution risk to SoftBank’s grand plan, as political will and initial commitments must translate into sustained financial backing, regulatory agility, and effective project management.

A salient example of this challenge is the flagship UK project by OpenAI to construct a major facility in north-east England. Despite being enthusiastically hailed by then-Prime Minister Sir Keir Starmer’s government in September, that initiative has since been placed on indefinite hold. This serves as a stark reminder that even the most high-profile partnerships, backed by powerful political endorsements, are subject to economic headwinds, technical challenges, and evolving corporate priorities, underscoring the inherent volatility in the nascent AI infrastructure market.

President Macron, since coming to power in 2017 on a decidedly pro-business platform, has consistently championed France as an attractive investment destination. His strategy includes the high-profile annual “Choose France” summit at the opulent Palace of Versailles, a carefully curated event designed to foster direct engagement between global CEOs and the French administration. This consistent emphasis on industrial and technological investment has created an environment conducive to mega-deals like SoftBank’s, demonstrating the efficacy of combining political will with economic incentives.

In recent years, the thematic focus of “Choose France” has increasingly pivoted towards AI and advanced technologies, reflecting a strategic alignment with global innovation trends. France has shrewdly leveraged its deep reliance on low-carbon nuclear electricity as a significant competitive advantage for attracting energy-intensive data centres. In an era of increasing ESG scrutiny and volatile energy prices, offering stable, relatively cheap, and environmentally friendlier power sources provides a compelling differentiator for investors seeking sustainable and cost-effective operational footprints.

The timing of this year’s “Choose France” event, and by extension, the SoftBank announcement, carries significant political weight. It occurs less than a year before France’s next highly uncertain presidential election, in which the far-right is currently polling ahead, and President Macron is constitutionally barred from seeking another term. This impending political transition introduces an element of long-term policy risk for such massive, multi-decade projects, as future administrations may have different priorities or regulatory approaches, potentially impacting the continuity and support for such foreign investments.

Market Impact: SoftBank’s colossal investment in French AI infrastructure sends a resounding signal across global capital markets: the AI revolution’s next frontier is foundational compute, and the race to own this critical utility is intensifying. For SoftBank, this marks a strategic doubling down on Masayoshi Son’s long-held vision, potentially yielding significant long-term returns but also entailing massive capital expenditure and execution risk. For France and Europe, it represents a crucial step towards digital sovereignty, attracting substantial foreign direct investment, and bolstering its position in the global tech landscape, though it will test the limits of their energy grids and regulatory agility. Investors in related sectors – from utilities and construction to semiconductor manufacturers and cooling technology providers – should anticipate sustained demand and potential growth opportunities. However, the cautionary tales of unrealised projects highlight the importance of diligent due diligence, as political shifts and market dynamics could introduce volatility into these capital-intensive, multi-decade ventures, underscoring that the promise of AI infrastructure is matched by its inherent complexities and financial demands.