Two of the world’s largest cloud computing companies are competing for the same prize: a lease inside a massive artificial intelligence (AI) data center that SoftBank is building in Ohio, according to reports from Nikkei Asia and Reuters. Google and Microsoft have each expressed interest in securing capacity at the facility, and according to those same reports, SoftBank could select tenants as early as spring 2026, although the timeline may shift if talks or regulatory steps drag on.
Neither Google, Microsoft, nor SoftBank has publicly confirmed the discussions. The reported bidding war, however, highlights just how scarce large-scale AI computing capacity has become and how aggressively the biggest players in tech are moving to lock it down.
Why Ohio, and why now
Ohio has emerged as one of the fastest-growing data center markets in the United States. The state offers comparatively affordable electricity, a package of tax incentives for large computing facilities, and a central geographic position that cuts latency for users across the eastern half of the country. Amazon Web Services, Google, and Meta already operate clusters of data centers in and around Columbus, Ohio.
The SoftBank facility appears to be connected to a broader U.S.-Japan investment framework that the White House formalized in September 2025. Under that agreement, Japan committed up to $550 billion in loans, guarantees, and other financial instruments directed at American energy, transportation, and digital infrastructure over multiple years. As of spring 2026, neither government has publicly disclosed how much of that $550 billion has actually been deployed or allocated to specific projects. The U.S. Department of Commerce has outlined a bilateral investment agenda that includes the Portsmouth initiative, which is an energy and infrastructure effort based in Portsmouth, Ohio. Media reports have linked SoftBank’s data center to that initiative, although the specific connection has not been confirmed by government agencies or by SoftBank itself.
If the connection holds, then it would mean that the facility is not simply a private real estate deal but a government-backed project channeling Japanese capital into American infrastructure at a moment when AI computing is straining the nation’s power grid and construction capacity.
The urgency of either Google or Microsoft securing capacity inside of SoftBank’s data center is notable because demand for AI training and inference capacity has outstripped what existing data centers can deliver. Alphabet, Google’s parent company, spent more than $75 billion on capital expenditures in 2025, with the bulk directed at data center construction, according to the company’s earnings disclosures. Similarly, Microsoft, which supplies computing power to OpenAI and runs its own Azure AI services, has committed tens of billions of dollars to expanding its global footprint. In a market where available capacity is measured in months of supply rather than years, a large block of space in a new, well-funded facility carries serious strategic weight.
SoftBank’s broader AI push
The Ohio project fits into a larger bet that SoftBank CEO Masayoshi Son has placed on AI infrastructure. In January 2025, Son appeared alongside President Trump to announce the Stargate project, which is a joint venture with OpenAI and Oracle that pledged up to $500 billion in AI infrastructure spending over four years. “This will be the most important project of this era,” Son said at the announcement, describing the scale of investment SoftBank planned to direct toward AI. SoftBank also controls Arm Holdings, the chip architecture company whose processor designs underpin much of the hardware running AI workloads worldwide.
Leasing capacity to Google and Microsoft would represent a different strategy than Stargate, which is tightly linked to OpenAI. An Ohio facility serving multiple cloud providers would position SoftBank as a landlord to the broader AI industry rather than a partner to any single company. That diversification could prove valuable if competitive dynamics shift or if any one tenant’s demand fluctuates, which is a possibility in a sector where spending cycles can turn quickly.
Critical details remain undisclosed
Many details regarding this project are still unknown. No public documentation describes the facility’s exact location within Ohio, its planned computing capacity, its power requirements, or its construction timeline. Additionally, it is unclear whether the site will sit adjacent to existing industrial infrastructure, share resources with power generation assets, or occupy newly developed land.
The bidding process is also somewhat unknown. The reports from Reuters describe an expected leasing process but do not specify whether companies beyond Google and Microsoft are also competing, what the lease terms look like, or how SoftBank plans to evaluate offers. Whether the facility will serve a single anchor tenant or multiple providers simultaneously has not been spelled out.
Permitting and power procurement could also shape the timeline of this data center project. Large-scale AI data centers can consume as much electricity as a small city. A single hyperscale facility can draw 100 megawatts or more, enough to power roughly 80,000 homes. There has been no public indication of which federal, state, or local permits have been secured by SoftBank, what environmental reviews are underway, or how SoftBank plans to source that power. These factors have slowed data center projects elsewhere in the country and could do the same here.
A Congressional Research Service analysis of the 2025 U.S.-Japan Framework Agreement notes that Congress is tracking oversight questions around the $550 billion commitment, including transparency in how projects are selected and how risk is shared between public and private actors.
What a lease deal would mean for the AI arms race
If Google or Microsoft signs a lease at the SoftBank facility, it would represent one of the first tangible outcomes of the U.S.-Japan investment framework in the AI sector. This outcome would also underscore a shift in how the largest technology companies compete: the race is no longer just about building better models but about controlling the physical infrastructure those models depend on.
For Ohio, a project of this scale could bring thousands of construction jobs and long-term technical positions, but it would also intensify pressure on the state’s electrical grid and water resources, which are concerns that communities near existing data center clusters have already raised.
The strongest reporting so far comes from Nikkei Asia’s sourcing, echoed by Reuters. According to those outlets, SoftBank could choose tenants as early as spring 2026, although the reports caution that the schedule depends on the pace of negotiations and regulatory approvals. Plans at this stage can still shift, and until official announcements follow, the Ohio project remains a high-stakes negotiation rather than a done deal.
