Just one day after announcing a monumental $100 billion investment deal with chipmaker Nvidia, OpenAI revealed new agreements with Oracle and SoftBank to build five additional data centers across the United States. This development further cements OpenAI’s role at the heart of the global artificial intelligence (AI) infrastructure boom, while raising questions about cost, scale, and long-term sustainability.

Expanding U.S. AI Infrastructure

OpenAI confirmed on Tuesday that the company will collaborate with SoftBank, the Japanese technology conglomerate, and Oracle, one of the largest cloud computing firms, to develop new high-capacity computing hubs in:

• Lordstown, Ohio
• Milam County, Texas
• Shackelford County, Texas
• Doña Ana County, New Mexico
• An additional site in the Midwest, which has not yet been publicly identified

These facilities represent the next stage of the Stargate Project, a $500 billion initiative announced in January at the White House alongside President Donald Trump. The ambitious plan involves building an extensive network of AI-focused infrastructure designed to accelerate research, product deployment, and the delivery of AI solutions to both businesses and consumers.

With the latest agreements in place, OpenAI has already secured over $400 billion in commitments for new data centers. Additional partnerships will be required to reach the project’s half-trillion-dollar target.

Role of Oracle and SoftBank

Under the terms of the deal, Oracle will finance and oversee the construction of three of the five newly announced sites. Once completed, OpenAI will purchase computing power directly from Oracle’s cloud infrastructure. Clay Magouyrk, Oracle’s co-chief executive, explained that the company intends to use a mix of innovative financing methods to cover construction costs, including partnerships with other investors, technology providers, and financial institutions.

“It’s really about bringing the right capital sources together, using new types of corporate structures and creative financing models,” Magouyrk said during the announcement.

SoftBank, meanwhile, will finance the two remaining data centers located in Ohio and Milam County, Texas. According to sources familiar with the arrangement, SoftBank plans to raise money through a combination of loans and bank financing. OpenAI will be directly involved in overseeing construction at these sites.

A Global Race for AI Power

The announcement comes amid a worldwide surge in demand for AI-driven infrastructure. Major technology companies—including Amazon, Google, Meta, Microsoft, and OpenAI itself—are collectively projected to spend more than $325 billion on data centers before the end of this year.

For OpenAI, the latest wave of investment follows a significant boost from Nvidia, which committed $100 billion to the startup just one day earlier. Nvidia has already injected an initial $10 billion into OpenAI, securing a roughly 2 percent stake in the company at a valuation of $500 billion. Over the coming years, Nvidia intends to make nine more $10 billion installments as OpenAI scales up its infrastructure.

This move underscores Nvidia’s growing influence as the dominant supplier of GPUs—the specialized chips that power AI systems like OpenAI’s ChatGPT and other large language models.

Microsoft, Nvidia, and the Web of AI Investment

OpenAI’s financial model has become a recurring story in the tech industry: raising money from the very companies it relies on for hardware, cloud services, and software tools. Between 2019 and 2023, Microsoft invested more than $13 billion in OpenAI, later generating billions in revenue by selling computing power back to the startup.

Nvidia’s new role follows a similar pattern. In exchange for its multibillion-dollar stake, OpenAI will rely heavily on Nvidia’s chips to power its rapidly expanding global data center footprint.

However, this cycle of heavy spending has raised concerns among analysts. While OpenAI currently generates billions in revenue, it is also spending tens of billions annually—primarily on computing power. If AI adoption slows or revenue streams fall short of expectations, the startup and its partners could face mounting financial risks.

The Scale of the Stargate Project

The Stargate Project represents one of the most ambitious technology infrastructure efforts in modern history. In Abilene, Texas, OpenAI and Oracle have already started construction on a large-scale facility that consumes roughly 1.4 gigawatts of electricity—enough to power hundreds of thousands of homes.

Reporters were given a glimpse of the site earlier this week, highlighting the sheer scale of the undertaking. According to project leaders, eight new data centers are planned under the Stargate banner, with two already operational and others progressing through various stages of development.

The energy requirements alone highlight one of the project’s key challenges: ensuring that these massive facilities can be powered sustainably while keeping operational costs manageable.

Expanding Beyond the U.S.

The company’s ambitions are not confined to North America. OpenAI has also announced plans for a data center in the United Arab Emirates (UAE) as part of a joint venture with Oracle, SoftBank, and the Emirati AI company G42. This expansion was made possible through a bilateral agreement between the Trump administration and the UAE, signaling the geopolitical importance of AI infrastructure as countries compete to secure their place in the AI economy.

By extending its footprint into the Middle East, OpenAI is positioning itself not only as a U.S. tech powerhouse but also as a global infrastructure player, capable of shaping the future of AI access across continents.

Industry Concerns and Legal Battles

Despite the optimism surrounding these announcements, industry experts continue to express skepticism about the sustainability of such aggressive expansion. Critics argue that the current pace of investment could saddle AI companies with massive debt if consumer and enterprise adoption of AI tools does not meet expectations.

Meanwhile, OpenAI continues to face legal challenges. In late 2023, The New York Times filed a lawsuit against OpenAI and Microsoft, alleging copyright infringement in relation to the training of AI models on news content. Both companies have denied wrongdoing, but the outcome of this legal dispute could shape how AI systems are built and trained in the future.

The Bigger Picture

OpenAI’s rapid expansion illustrates the larger trend sweeping through the global technology sector: the race to control the infrastructure of artificial intelligence. As AI tools become more integrated into everyday business operations and consumer products, control over the data centers powering these systems will determine which companies dominate the next generation of digital innovation.

For now, OpenAI’s strategy appears clear—secure vast sums of capital, form alliances with technology giants like Nvidia, Oracle, and SoftBank, and build out infrastructure at an unprecedented pace. Whether this approach will prove financially sustainable remains to be seen.

What is certain, however, is that the AI revolution is not just about algorithms or software—it is about the immense physical and financial machinery required to keep those algorithms running at global scale. OpenAI’s latest announcement is a reminder that the future of AI will be shaped as much by steel, silicon, and electricity as by lines of code.