Inside the Multi-Trillion-Dollar Race to Build AI Infrastructure: Microsoft, Oracle, Meta and More
Tech giants are pouring unprecedented sums into AI infrastructure, sparking a new gold rush in data centers and energy. Nvidia CEO Jensen Huang recently projected that between $3 trillion and $4 trillion will be spent on AI infrastructure by 2030, with much of that coming from companies racing to deploy large AI models. This explosion in spending is stretching power grids, driving up energy demand and testing the limits of global construction capacity.
Microsoft’s early bet on OpenAI helped set the stage for today’s AI boom. Beginning with a $1 billion investment in 2019, Microsoft became OpenAI’s exclusive cloud partner, offering Azure credits to cover massive training costs. That stake has since grown to nearly $14 billion. Although OpenAI has since diversified its providers, the Microsoft–OpenAI model is now the blueprint for cloud-AI partnerships. Amazon-backed Anthropic and Google Cloud’s deals with emerging AI firms echo this same approach. Nvidia’s recent $100 billion investment into OpenAI underscores just how high the stakes have become.
Oracle has emerged as an unexpected heavyweight in the AI infrastructure race. In June 2025, it signed a $30 billion cloud services deal with OpenAI—larger than its entire previous-year cloud revenue. By September, it followed with a record-shattering five-year, $300 billion compute power agreement beginning in 2027. Those deals sent Oracle’s stock soaring and briefly made founder Larry Ellison the richest person in the world, cementing Oracle’s status as a major force in the AI data center market.
Meta, Elon Musk’s xAI, and others are building massive hyperscale facilities with real-world environmental implications. Meta alone plans to spend $600 billion on U.S. infrastructure by 2028, including two mega data centers: Hyperion in Louisiana (powered partly by nuclear energy) and Prometheus in Ohio (powered by natural gas). Meanwhile, xAI’s hybrid data center in Tennessee has drawn criticism for air pollution linked to its natural gas turbines, highlighting the environmental trade-offs of the AI boom.
The “Stargate” project shows both the ambition and the risks of such mega-deals. Announced by President Trump in 2025 as a $500 billion joint venture between SoftBank, OpenAI and Oracle, Stargate was billed as the largest AI infrastructure project ever. But despite initial hype, reports suggest the partners are struggling to reach consensus. Still, construction is underway on eight Texas data centers, a sign that even amid uncertainty, the race to build AI’s backbone is accelerating.
source: techcrunch