SoftBank Rides the A.I. Infrastructure Boom

SoftBank’s Ascent Reflects a New Center of Gravity in the AI Boom

SoftBank has overtaken Toyota to become Japan’s most valuable listed company, a striking market milestone that coincided with an even larger statement about where the artificial-intelligence race is heading: not just into software and chatbots, but into the physical infrastructure needed to power them.

The Japanese technology conglomerate’s market value climbed to roughly ¥48.8 trillion on June 1, surpassing Toyota’s about ¥45.9 trillion after a sharp rally in its shares. At nearly the same moment, Masayoshi Son, SoftBank’s founder and chief executive, leaned into a characteristically grand vision of the industry’s future, saying the A.I. revolution would be “50 times bigger” than the dot-com boom.

SoftBank also announced one of Europe’s most ambitious A.I. infrastructure plans to date: an investment of up to €75 billion, or about $81 billion, to develop 5 gigawatts of A.I. data-center capacity in France. An initial €45 billion phase is aimed at building 3.1 gigawatts in the northern Hauts-de-France region.

Taken together, the developments show how investor excitement around A.I. is increasingly migrating from the applications layer to the harder, costlier foundations beneath it — electricity, land, high-end chips, cooling systems and grid access. In that sense, SoftBank’s rise in value is not simply a bet on digital intelligence, but on the industrial machinery required to sustain it.

A Bet on Compute, Power and Scarcity

For months, global markets have rewarded companies tied to A.I. models, semiconductors and cloud demand. But as the frenzy matures, a new constraint has become clearer: training and running advanced systems requires enormous amounts of compute, and compute requires enormous amounts of power.

That reality helps explain why SoftBank’s French move has drawn such attention. Five gigawatts is an immense target, large enough to place the project among the biggest announced A.I. data-center expansions in Europe. SoftBank has said the facilities are intended to serve demand from A.I. companies, cloud providers, enterprises, public institutions and researchers.

France has been marketing itself as a logical host for such projects. The country offers relatively abundant low-carbon electricity, a feature that is increasingly valuable as technology companies come under pressure to secure power at scale without sharply worsening their emissions footprint. Industrial land with favorable grid connections — once a niche consideration — has become a strategic asset in the global contest to host A.I. infrastructure.

The investment announcement came around the government-backed Choose France summit, an event meant to attract foreign capital and showcase France as a destination for major industrial projects. The timing underscored a broader shift in Europe, where governments are trying to turn concerns about technological dependence into an industrial policy push of their own.

Why Investors Are Rewarding SoftBank

SoftBank’s climb past Toyota also says something about how markets are valuing A.I. exposure. The company has increasingly been seen as a vehicle for the sector, with recent profit growth helped by gains tied to A.I.-related holdings, including OpenAI and Arm, the British chip designer in which SoftBank holds a controlling stake.

Arm has become a central piece of the A.I. investment story because its chip architecture sits at the heart of vast swaths of modern computing, from smartphones to data-center systems. Optimism surrounding Arm, as well as broader enthusiasm about OpenAI-linked opportunities, has helped fuel SoftBank’s recent rally.

That marks a sharp contrast with SoftBank’s more uneven recent history. The group spent years oscillating between triumph and retrenchment as lofty technology bets rose and fell. Now, with A.I. commanding global capital and political attention, Mr. Son has recast SoftBank as a pivotal participant in what he argues will be a transformational technological shift.

His rhetoric has often been expansive, and Monday’s comments fit that pattern. But the market appears willing, at least for now, to accept the scale of his ambition — especially when it is accompanied by concrete infrastructure plans in regions eager to host them.

France’s Broader Industrial Pitch

SoftBank is not entering France from scratch. Its announced footprint there already includes a 1-gigawatt campus in Bosquel with the partner Sesterce, an industrial manufacturing cluster in Dunkirk with Schneider Electric, and its selection by EDF as preferred bidder for a 400-megawatt data-center project at the former Bouchain power plant site.

Those pieces suggest a strategy broader than simply leasing server halls. The buildout appears aimed at creating an ecosystem of energy, industrial support and high-performance computing capacity — the sort of vertically connected platform that many governments and technology companies now see as necessary for A.I. leadership.

For France, the appeal is obvious. Hosting A.I. infrastructure promises investment, construction, industrial demand and a stronger role in a field long dominated by American and, increasingly, Chinese players. For Europe more broadly, such projects are part of a push to ensure that the continent does not remain merely a regulator of A.I. while the most important computing clusters are built elsewhere.

The Hard Part Comes Next

Yet the gap between announcing tens of billions of euros and delivering operating capacity remains large.

SoftBank’s commitment is framed as “up to” €75 billion, and the full 5 gigawatts will depend on financing, permits, environmental reviews, construction schedules, equipment availability, transmission access and customer demand. It is also unclear how much of the planned capacity already has anchor tenants committed to use it, a key issue in a market where demand forecasts are soaring but still difficult to pin down.

That uncertainty is especially relevant in Europe, where leaders have become more assertive in courting A.I. investment but still face questions about speed. Can headline pledges be converted into live computing clusters quickly enough to narrow the gap with the United States and China? Can power and grid access be secured without delays that undermine competitiveness? And can the economics of these massive facilities hold up if demand proves more uneven than today’s exuberance suggests?

For now, investors are signaling that they believe the answer may be yes — or at least that the scramble for A.I. capacity is only beginning.

SoftBank’s new standing atop Japan’s market rankings offers a vivid measure of that belief. In an earlier technology era, valuations often hinged on platforms, users and software narratives. In this one, they are increasingly being shaped by who can secure the turbines, substations, chips and concrete needed to turn A.I. ambition into industrial reality.