Who Finances Masayoshi Son’s Next Server?

SoftBank’s $4 billion takeover of DigitalBridge would give Masayoshi Son’s AI ambitions a data-center investment group with a role beyond direct corporate funding. DigitalBridge CEO Marc Ganzi told the Financial Times that the group would become SoftBank’s “third-party infrastructure arm.” His description concerns the intended role after the takeover, not evidence that every future server has secured financing.

The available Financial Times summary does not establish whether the transaction has closed, how SoftBank will finance the purchase, or what project obligations SoftBank will retain. The full report was unavailable for review. The $4 billion acquisition price therefore cannot be treated as the infrastructure budget, and the third-party label cannot establish that SoftBank has moved liabilities off its balance sheet.

Granted, the strongest case is substantial. Outside investors can finance infrastructure beyond the resources of one corporate owner. A fund manager can assemble capital from multiple investors and spread exposure across projects. If investors accept clearly defined risks, SoftBank could pursue more capacity without supplying every construction dollar itself. That is a financing advantage, not merely an exercise in vocabulary.

The capital source is not the demand forecast

Still, the physical investment needs customers. Data centers require paying users, sufficient utilization, and revenue that covers operating costs and financing. Raising money can accelerate construction without establishing those conditions. The ambition is in-house; the capital gets a guest invitation. Neither arrangement guarantees that the guest receives an adequate return.

Ownership and funding also need separate columns. SoftBank’s ownership of DigitalBridge would not mean that SoftBank owns every asset managed by DigitalBridge. Managing a fund does not make the manager the supplier of all its capital. Project debt creates claims for lenders; guarantees can place obligations on a separate sponsor. These distinctions explain why balance-sheet relief remains a hypothesis until the relevant terms are visible.

Yet outside capital can impose useful discipline. Investors and lenders may demand credible customers, limits on borrowing, and protection against poorly priced construction commitments. The competing concern is whether a manager under SoftBank ownership would face pressure to favor Son’s expansion plans over investor returns. That is an incentive to examine, not a conflict already proved. Independent investment decisions and transparent related-party terms would strengthen the favorable case.

Near-term fundraising would demonstrate financing capacity. Longer-term returns would demonstrate whether the infrastructure earns its cost. The decisive disclosures are customer contracts, project debt terms, capital commitments, and any SoftBank guarantees. If independent customers support revenues and outside investors bear clearly bounded losses, the third-party model genuinely shares risk. If SoftBank guarantees payments or absorbs shortfalls, the infrastructure arm may extend its reach without reducing its exposure.