Big Tech's $635B AI Spending Faces Energy Shock Test

S&P Global warns that Big Tech’s massive AI infrastructure investments may require revision. Microsoft, Amazon, Alphabet, and Meta planned to spend approximately $635 billion on AI infrastructure in 2026, up from $383 billion the previous year. But persistently high energy prices threaten the entire equation.

The Hormuz crisis has pushed Brent crude above $100 per barrel since early March, peaking at $126. As of March 30, Brent trades near $106 and WTI around $100. The IEA has authorized its largest-ever emergency oil stock release: 400 million barrels.

AI data centers are extremely power-intensive. Projections indicate data center consumption could reach 1,050 terawatt-hours in 2026. In the US alone, AI-specific servers may consume up to 326 TWh annually by 2028, representing roughly 12 percent of total national power demand.

Combined with much of the new capacity running on fossil fuel power plants, this creates a dual challenge: higher costs and increased carbon footprint. Analysts suggest sustained oil prices above $100 could force tech giants to cut or delay expansion plans.

For CIOs, this means potentially higher cloud prices and longer wait times for GPU capacity through 2026 and 2027.