SoftBank Group said on December 30 (China time) it will acquire U.S. digital infrastructure investor DigitalBridge for $4 billion in cash at $16 per share. The deal, framed as a push to build next‑generation AI data centers, is expected to close in the second half of 2026 pending approvals.
This article aggregates reporting from 3 news sources. The TL;DR is AI-generated from original reporting. Race to AGI's analysis provides editorial context on implications for AGI development.
SoftBank’s move to buy DigitalBridge is a classic “picks and shovels” bet on the AI gold rush. Rather than chasing model labs directly, Masayoshi Son is locking up the underlying infrastructure—data centers, fiber, towers and edge sites—that every frontier model ultimately depends on. DigitalBridge already manages over $100 billion in digital assets, giving SoftBank an immediate footprint in the facilities required for multi‑gigawatt AI super‑clusters and latency‑sensitive inference workloads.
Strategically, this folds neatly into Project Stargate and SoftBank’s ambition to be an “ASI platform” company rather than just a financial investor. As hyperscalers and frontier labs race toward larger, always‑on models, control of power‑dense, well‑connected campuses becomes a chokepoint. By owning both capital and concrete, SoftBank positions itself as a gatekeeper for future OpenAI‑, Anthropic‑ or xAI‑scale deployments. That intensifies competition with traditional infra players and with cloud hyperscalers building their own real estate, and it will likely push rivals to pursue similar roll‑ups or long‑term leasing deals.
For the broader AI ecosystem, more capital flowing into purpose‑built AI data centers should ease some near‑term compute scarcity—but it also concentrates leverage in a small set of infra landlords.