The short version

SoftBank Group said on December 29, 2025 that it will acquire DigitalBridge Group for about $4 billion enterprise value, or $16.00 per share in cash—a 15% premium to DigitalBridge’s December 26 close—pending regulatory approvals with closing targeted for the second half of 2026. DigitalBridge will continue operating as a separately managed platform led by CEO Marc Ganzi. SoftBank | DigitalBridge | WSJ

15%
Deal premium (to Dec 26 close)Source: SoftBank press release, Dec 29, 2025
$108B
DigitalBridge assets under managementSource: DigitalBridge Q3 2025 results
Conceptual editorial image showing a modern data center aisle with glowing fiber links connecting the U.S. and Japan, symbolizing SoftBank and DigitalBridge’s global infrastructure footprint

Why this deal matters for AI infrastructure

SoftBank’s purchase is not a bet on one AI model or app—it’s a bet on the physical substrate of AI: power, land, compute, fiber, and cooling. DigitalBridge is one of the largest dedicated investors in that substrate, with positions across hyperscale data centers, towers, and long‑haul fiber. That makes the company a natural puzzle piece for SoftBank’s Stargate initiative with OpenAI and Oracle, which aims to mobilize hundreds of billions of dollars to build multi‑gigawatt AI data centers in the U.S. OpenAI | SoftBank | Financial Times

What, exactly, is SoftBank buying?

  • A $108B‑AUM digital infrastructure specialist that owns or manages stakes in:
  • A team and playbook oriented around sourcing land, power, and interconnects—precisely the capabilities needed to turn AI capital into usable, reliable compute capacity.

For readers tracking the sector’s consolidation: this deal lands as hyperscaler demand and AI training loads pull forward multi‑year data center pipelines. The transaction follows DigitalBridge’s 2024–2025 push to scale Vantage globally and acquire Yondr, both centered on AI‑ready campuses. DigitalBridge | Yondr

How the pieces fit: from chips to cables to clean power

SoftBank’s thesis has become clear in 2025: control or influence the full AI stack—from Arm‑based compute design (via Arm and the acquisition of Ampere Computing) to hyperscale data centers (Stargate, Oracle partnership) to the energy and grid interface. DigitalBridge slots into the middle of that stack.

  • Stargate momentum: OpenAI and partners laid out a path to 10+ GW of U.S. AI data center capacity over the next few years, with multi‑site expansions across Texas, New Mexico, and Ohio. OpenAI | SoftBank
  • Power and siting: SoftBank‑affiliated SB Energy is developing large, power‑rich campuses (for example in Milam County, Texas) and aims to deliver multi‑gigawatts of data center capacity starting in 2026. DigitalBridge’s build‑operate‑partner model can shorten time‑to‑compute by pairing those sites with proven developers and operators. SB Energy
  • Compute roadmap: By pairing Ampere’s Arm‑based CPUs with Nvidia‑class accelerators (through partners) in DigitalBridge‑scale campuses, SoftBank is seeking cost‑efficient, power‑dense designs that can be replicated quickly.

Deal snapshot

Key terms of SoftBank–DigitalBridge

TermDetail
Headline valueApproximately $4.0B enterprise value (including debt)
Consideration$16.00 per share in cash
Premium15% to Dec 26, 2025 close; ~50% to unaffected 52‑week average
StructureDigitalBridge to operate as a separately managed platform under CEO Marc Ganzi
Approvals & timingCustomary regulatory approvals; expected close in H2 2026
Strategic intentAccelerate build‑out of next‑gen AI data centers and connectivity for Stargate‑scale compute

Sources: SoftBank press release, DigitalBridge release, WSJ

What it means for builders, CIOs, and automation leaders

  • Faster paths to capacity: If SoftBank leverages DigitalBridge’s hyperscale partners (Vantage, Yondr) and SB Energy’s power pipelines, it could shorten lead times from “capital committed” to “compute online.” That helps teams stuck on GPU waitlists or paying premiums for scarce capacity. Vantage | SB Energy
  • More options for multi‑cloud AI: Oracle’s deepening role in Stargate, alongside OpenAI’s continued use of Azure, suggests a more heterogeneous AI infrastructure landscape. Enterprises may benefit from diverse regions, architectures, and pricing models. OpenAI
  • Productivity knock‑on effects: Downward pressure on training/inference costs improves unit economics for AI‑enabled automation—freeing budgets to move proofs‑of‑concept into production, and to push compute to the edge via fiber‑rich, low‑latency sites such as AtlasEdge in Europe. AtlasEdge
TipActionable next steps for enterprise AI teams
  • If you plan major 2026–2028 AI deployments, start power‑and‑site diligence now; grid interconnects often outlast chip lead times.
  • Ask providers for transparent roadmaps on where and when new capacity comes online (and how it’s powered). Tie commitments to staged SLAs.
  • Design models and pipelines to be portable across clouds and colocation partners. The SoftBank–DigitalBridge tie‑up could create new regions and economics worth arbitraging.

Risks and open questions

  • Regulatory clearance: The companies cite “customary approvals.” Given DigitalBridge’s U.S. critical‑infrastructure footprint and SoftBank’s ownership, national‑security review will be in focus. SoftBank
  • LP alignment: DigitalBridge manages third‑party capital. Maintaining strict fiduciary and conflict‑management frameworks will be essential so fund investors benefit alongside SoftBank’s strategic goals. DigitalBridge
  • Power is the moat: Even with capital and developers, interconnection queues and transmission constraints remain the gating factor for AI capacity. SB Energy and peers are racing to line up multi‑GW sites, but timelines can slip due to permitting or grid upgrades. SB Energy

Market reaction and timeline

DigitalBridge shares jumped on the news and had spiked earlier amid reports of a pending deal; the offer equates to a 15% premium versus the prior close. The companies guide to a second‑half 2026 completion, suggesting a lengthy approval cycle consistent with the size and sensitivity of the assets. Barron’s | Reuters

Bottom line

SoftBank isn’t just buying an asset manager—it’s buying time. DigitalBridge’s relationships, sites, and operating know‑how could help convert SoftBank’s aggressive AI capital plan into real, rentable compute faster than rivals. For teams building automation and AI products, more capacity—delivered sooner and closer to users—translates into lower latency, lower costs, and the ability to scale beyond pilots.

Sources