Five Hyperscalers Spent $1.1 Trillion as U.S. Adults Oppose Local AI Data Centers

Five hyperscalers spent $1.1 trillion on capital expenditures over five years, while 71% of U.S. adults oppose an AI data center in their area, underscoring public resistance to the costly buildout.

Amazon, Alphabet, Microsoft, Meta and Oracle accounted for the five-year spending total. Sell-side analysts project another $5.3 trillion in capital spending through 2030. That projection is not a commitment from the companies.

Public companies whose earnings-call language about AI was more positive accounted for all growth in public-company capital investment since the first quarter of 2024. In the first nine months of 2025, their investment growth exceeded that of other public firms by about 25 percentage points. The figures cover publicly traded companies, not private-market or venture-capital investment.

A Feb. 17-23, 2026, survey found that 63% of U.S. adults thought AI was moving too fast, 19% considered its pace about right and 2% thought it was moving too slowly. A separate March 2-18 survey found that 71% opposed local construction of an AI data center; 48% were strongly opposed, while 27% favored it. That poll included 1,000 adults and had a margin of sampling error of ±4 percentage points.

At June 30, 2026, CoreWeave’s total liabilities stood at $72.046 billion. Separate figures for the same date put its total indebtedness at $35.6 billion and operating lease liabilities at $16.3 billion. Its bonds have faced market pressure.