Google Takes Direct Control of a US$17 Billion Wyoming Data-Centre Project

Google acquired Crusoe’s pre-developed data-centre site in southeast Wyoming on November 6, one day after announcing a long-term US$17 billion investment in its first data centre in the state. Crusoe had selected the site, secured governmental approvals and developed the power strategy; the purchase price and closing terms were not disclosed.

The handoff matters because it moves Google from prospective user to owner of the entitled site. The largest AI buyers are no longer choosing only between locations. They are deciding how much development risk to internalise when power, approvals, equipment and construction schedules can determine whether billions of dollars of computing capacity arrive on time. Google gains direct control over that delivery chain. Crusoe exits the Wyoming pre-development role, although it continues to develop another Google campus in Texas.

Power is the centre of the location proposition. Google says the Wyoming project is contracted with American Electric Power for a planned 702 MW fuel-cell development using natural gas and carbon capture and storage. It will also source power from gas generation operated by Tallgrass Energy and Black Hills Energy. The company says it will pay the full cost of the power it uses and the infrastructure directly caused by its growth. Air cooling is intended to reduce water demand.

Those commitments answer the questions now defining data-centre approvals: who pays for grid expansion, what happens to household rates, how much water the project consumes and whether low-carbon claims survive scrutiny. They do not settle them. The fuel cells, capture infrastructure and sequestration pathway are planned rather than operating; Google has not published a year-by-year spending schedule, computing capacity or construction completion date.

The immediate winners are Google, which controls a de-risked site; utilities and energy partners with long-term demand; and Wyoming, if the project reaches construction at the announced scale. Independent developers face a tougher comparison. Their value lies in assembling land, permits and power quickly, but hyperscalers may take ownership when cost and timetable control become more valuable than an arm’s-length lease. Locations that rely on an intermediary’s headline pipeline should therefore distinguish contracted capacity from speculative capacity.

For investment agencies, the lesson is to diligence the delivery structure, not simply celebrate the capital number. A decision-ready data-centre file should identify the landowner, end user, developer, power provider, interconnection schedule, water model, permitting status and party carrying each infrastructure cost. Agencies should also test whether announced carbon capture, generation and grid investments have financing, approvals and executable schedules. A memorandum or power concept is not capacity.

The community offer has become part of site readiness. Google expects hundreds of millions of dollars in local property taxes and has created a council with its project partners after residents asked for clearer economic and workforce benefits. That is useful, but the measure will be transparent agreements and delivery: tax revenue, skilled local employment, infrastructure costs and enforceable protections for other users of power and water.

What to watch: confirmation of the acquired site’s capacity and construction schedule, regulatory treatment of the power package, and whether Google publishes milestones against the US$17 billion commitment. Competing locations should assume that the winning bid now requires an investable energy system and a credible community bargain—not merely available land and an incentive package.