Big Tech hyperscalers are on pace to invest $700 billion in AI infrastructure in 2026, marking an unprecedented capital sprint that reshapes how the world’s largest technology companies are allocating resources. Amazon, Alphabet, Microsoft, and Meta are collectively committing to the historic spending level, with the overwhelming majority directed toward data centers, chips, and networking equipment needed to power advanced AI systems.
The scale of this investment dwarfs previous technology buildouts. Amazon leads individual company spending at approximately $200 billion in 2026 capex, followed by Alphabet at $175 billion to $185 billion, Microsoft at $120 billion to $190 billion, and Meta at $115 billion to $135 billion, according to multiple analyst estimates and company guidance.
This spending surge represents a dramatic acceleration from prior years. Capital expenditures among the four hyperscalers reached $427 billion in 2025, meaning 2026 spending is projected to nearly double year-over-year. Goldman Sachs reported that consensus estimates for capex growth had implied roughly 20 percent increases at the start of both 2024 and 2025, yet actual growth exceeded 50 percent in both years—a pattern that has continued into 2026.
The urgency behind the buildout stems from intense competition to secure AI compute capacity. Companies are racing to deploy data center infrastructure capable of handling the extraordinary power and cooling demands of modern large language models and other AI workloads. As AI workloads push power density higher, the cost to construct data centers in the AI era has risen significantly, according to Goldman Sachs analysis published in May 2026.
Analysts project the spending trajectory to accelerate further. Goldman Sachs’ baseline model implies $765 billion in annual AI capital expenditure in 2026, growing to $1.6 trillion in annual capex by 2031. Fortune reported in late June 2026 that analysts forecast continued growth, with capital expenditures reaching $650 billion in 2026 and surpassing $1.1 trillion in 2027.
The $700 billion figure reflects what some analysts describe as a once-in-corporate-history infrastructure buildout. Hyperscalers are utilizing debt financing to fund the massive data center expansion, raising investor scrutiny about long-term return on investment. A top Wall Street analyst at D.A. Davidson, Gil Luria, framed the debate over AI capital spending as a timing problem, noting that when companies build a data center, it is already pre-sold, suggesting underlying demand justifies the expenditure.
Sources
- Intellectia AI — hyperscaler capex projections and year-over-year growth rates for 2026
- Yahoo Finance — Big Tech capex spending forecasts and component breakdown (chips, servers, data centers)
- The Futurum Group — AI capex plans for Microsoft, Amazon, and Alphabet with $700B aggregate spending
- Fortune — analyst forecasts for 2026 and 2027 capex, debt financing trends, and ROI context
- Goldman Sachs — baseline AI CapEx model ($765B in 2026, $1.6T by 2031), historical capex growth rates (50%+ in 2024–2025), and data center cost increases
- RBC Wealth Management — capex growth trajectory ($427B in 2025) and momentum assessment
- 24/7 Wall St. — analyst commentary from D.A. Davidson on AI spending returns and pre-sold capacity
