On January 3, 2025 Microsoft vice chair and president Brad Smith published a post titled “The Golden Opportunity for American AI” in which he put a hard number on the company’s AI infrastructure plans. In his words: “In FY 2025, Microsoft is on track to invest approximately 80 billion dollars to build out AI-enabled datacenters to train AI models and deploy AI and cloud-based applications around the world. More than half of this total investment will be in the United States, reflecting our commitment to this country and our confidence in the American economy.”
The figure was striking for its scale and for being stated so plainly by the company itself. Eighty billion dollars in a single fiscal year, for data centers alone, marked a step-change from prior norms of corporate capital spending and set a reference point that competitors were measured against. The post framed the spending as part of a national AI strategy, pairing infrastructure investment with arguments about workforce skills and US competitiveness.
This commitment is one anchor of the broader AI capex supercycle. Coming at the start of 2025, it signaled that the largest cloud companies intended to keep accelerating rather than pause - and it preceded a year in which combined hyperscaler capital spending climbed well past 200 billion dollars.
Why a business reader should care: this is a first-party statement of intent from one of the world’s largest companies, quantifying how much it would spend building the physical backbone of AI in a single year. It is a concrete measure of the conviction - and the financial risk - behind the AI buildout.