OpenAI’s Stargate data center project is running into a harder financing environment, according to Bloomberg. The project is designed as a major AI infrastructure effort in the U.S. and overseas, but tariff-related economic uncertainty is now weighing on its progress.
The stakes are large because Stargate aims to raise up to $500 billion for AI infrastructure. That ambition depends not only on demand for computing capacity, but also on investor confidence, predictable construction costs, and a financing structure that can support large data center buildouts.
Why Stargate Is Facing New Friction
Bloomberg reports that banks, private equity investors, and asset managers have become wary of investing in Stargate. The caution is tied to several pressures arriving at the same time: growing market volatility, cheaper AI services, and the possibility that tariffs could make data center construction more expensive.
For a project centered on AI infrastructure, the cost of physical buildouts matters. Data centers depend on many categories of hardware and systems, and tariff increases can affect the economics of the parts needed to construct and operate them.
That uncertainty can make backers slower to commit. Even when investors are interested in AI, they still need a clear view of project costs, likely returns, and the financing structure behind the plan. If the cost base is shifting, the risk calculation changes.
The Cost Question Is Central
Tariffs could raise the price of several inputs used in data center buildouts. Bloomberg cited an analysis by TD Cowen that pointed to higher prices for server racks, cooling systems, chips, and other components.
According to that analysis, these increases could contribute to overall build cost rises of 5-15% on average. For Stargate, that range matters because the project’s target is up to $500 billion. Even a moderate increase in construction costs can affect how investors evaluate timing, financing, and scale.
The issue is not only whether costs rise, but whether they are predictable enough for investors to model. Data center infrastructure requires large commitments before revenue is realized. When important components become more expensive, or when investors believe prices may keep moving, financing discussions can become more difficult.
SoftBank Has Not Yet Set the Financing Path
SoftBank is a key part of the current uncertainty around Stargate. In January, SoftBank said it would contribute significant capital to the project. But Bloomberg reports that SoftBank has not yet developed a financing template or begun detailed discussions with potential backers.
That detail is important because a project of this size needs more than broad interest. It needs a structure that tells potential investors how capital will be raised, deployed, and managed. Without that template, backers may have fewer concrete terms to evaluate.
The reported delay does not mean the project has been abandoned. It does show that Stargate is still working through basic financing questions while the broader market becomes more cautious about AI infrastructure spending.
Investor Concern Extends Beyond Tariffs
Tariff-related cost pressure is only one part of the picture. Bloomberg also reports that investors are growing wary of an overcapacity spike. That concern reflects the possibility that data center supply could expand faster than demand justifies.
The caution comes as some major technology companies reassess their own data center strategies. Bloomberg notes that Microsoft and Amazon have adjusted their approaches and, in some cases, pulled back on construction projects.
For Stargate, that broader shift matters. If large technology companies are becoming more selective about building, investors may look more carefully at any new infrastructure plan, even one led by OpenAI. The question becomes not just whether AI will need more computing power, but how much capacity should be built, where it should be built, and how quickly.
What This Means for AI Infrastructure
Stargate illustrates a basic tension in the AI market. The industry’s infrastructure ambitions are large, but the physical systems behind those ambitions remain exposed to financing conditions, construction costs, supply chains, and investor sentiment.
Cheaper AI services add another layer to that tension. If AI services become less expensive, investors may ask whether the economics of new infrastructure projects still support the same level of capital spending. That does not remove the need for data centers, but it can change the pace and terms of investment.
The reported slowdown around Stargate therefore points to a practical reality: AI infrastructure is not only a technology story. It is also a capital allocation story. Building the systems behind AI requires confidence from financial institutions, private investors, and asset managers, especially when the target is up to $500 billion.
For now, Bloomberg’s report suggests that tariff uncertainty, rising buildout costs, financing questions at SoftBank, and concern about overcapacity are all creating friction for OpenAI’s Stargate project. The ambition remains large, but the path to funding and construction appears more complicated than the headline number alone suggests.