- OpenAI and SoftBank Group each commit $500 million to SB Energy to build multi‑gigawatt AI‑focused data centres.
- The move underscores growing energy demands and strategic infrastructure competition in the AI sector, with implications for costs, jobs and local communities.
What happened：$1bn Investment in AI Infrastructure Secures Future Power Supply
OpenAI and SoftBank Group Corp. have jointly invested a combined $1 billion in SB Energy, a SoftBank‑backed energy and infrastructure company, to accelerate the development of large‑scale data centres and supporting energy assets in the United States. Each organisation is contributing $500 million in equity financing, marking a significant injection of capital directed at the physical infrastructure needed to support next‑generation artificial intelligence workloads.
Under the partnership, SB Energy has been selected to build and operate a previously announced 1.2‑gigawatt data centre in Milam County, Texas, which is intended to supply the huge amounts of power required for AI compute tasks.SB Energy is also developing several other multi‑gigawatt data centre campuses, with initial facilities expected to enter service later in 2026.
The investment is being framed by the companies as part of Stargate, a broader AI infrastructure platform that OpenAI, SoftBank and partners unveiled with plans for multibillion‑dollar build‑outs. However, critics have raised concerns about the scale and opacity of these plans, particularly given intense competition from cloud providers and mounting questions about cost and energy consumption.
Why it’s important
This deal highlights a critical shift in the AI industry: the recognition that advanced artificial intelligence is not just a matter of software and models, but also of energy‑intensive hardware and infrastructure. As AI models grow larger and more complex, the demand for data centre capacity — and the electricity to run it — has surged, prompting strategic investments like this one.
The emphasis on energy and physical build‑outs reflects a broader arms race among technology firms to secure long‑term compute capacity, with rivals such as Google, Meta and Microsoft making their own infrastructure commitments.
Yet despite the enthusiasm, questions remain. The sheer scale of spending required to sustain these efforts — potentially in the hundreds of billions — raises concerns about economic viability, environmental impact, and equitable distribution of benefits. SB Energy’s plans to create jobs and support local grids will be closely watched, but analysts note that projects of this size can also place strain on regional resources and demand careful oversight.
By tying AI’s future to massive physical infrastructure, the industry may be entering a new phase — one where power generation and data‑handling capacity are as decisive as the models themselves.
Also read: Former OpenAI researcher revises AGI timeline, pushing superintelligence hopes into the 2030s
Also read: SoftBank’s $4bn DigitalBridge buyout signals deepening AI infrastructure play
