Big Tech AI Spending Hits $725B as Infrastructure Race Intensifies
The four largest U.S. technology companies plan to spend a combined $725 billion on artificial intelligence infrastructure in 2026, The Decoder reported, in a record capital expenditure push on data centers, semiconductors and AI systems.
Google, Amazon, Microsoft and Meta have committed the record sum to data centers, semiconductors and related AI infrastructure for 2026, as reported by The Decoder, citing Financial Times figures. The total represents a substantial increase in capital expenditure as each company races to secure the computing power needed to train and deploy increasingly sophisticated AI systems.
The spending commitments — disclosed through a combination of earnings calls, investor filings and corporate announcements — reflect a broad industry consensus that AI infrastructure demands will continue to grow. All four companies have signaled in recent quarters that they see no near-term ceiling on AI-related capital investment.
The scale of the outlay has direct implications for the U.S. economy. The bulk of planned data center construction is concentrated domestically, driving demand for American semiconductor suppliers, construction firms, electrical utilities and skilled labor. Chip manufacturers including Nvidia, AMD and Intel stand to benefit from sustained procurement at this level.
Each of the four companies has staked its competitive strategy on AI dominance. Microsoft has deepened its partnership with OpenAI while building out Azure AI infrastructure. Google has invested heavily in its Gemini model family and custom tensor processing units. Amazon has expanded AWS AI services and its custom Trainium chips. Meta has committed billions to open-source AI development and the infrastructure to support it.
The $725 billion figure dwarfs combined tech capital spending from just three years ago, when the four companies’ total capex hovered closer to $150 billion annually. The rapid acceleration has raised questions among some analysts about whether returns on AI investment will materialize quickly enough to justify the pace of spending.
Despite periodic investor concerns about the payoff timeline, all four companies have maintained or increased their spending guidance in recent earnings reports, suggesting corporate leadership sees AI infrastructure as a generational investment rather than a short-term bet.
The spending race also has geopolitical dimensions. U.S. export controls on advanced semiconductors to China have made domestic AI infrastructure buildout a matter of strategic importance, and the concentration of investment by American companies reinforces the United States’ position as the center of gravity for AI development.
