Big Tech AI spending hits $1.12 trillion in 2027 projection
Jefferies projects major hyperscalers will spend over a trillion dollars on AI infrastructure by 2027, despite rising debt.
The artificial intelligence industry is entering its largest investment cycle in technology history. Jefferies estimates that major hyperscalers, including Microsoft, Amazon, Alphabet, Meta, and Oracle, will spend approximately $1.12 trillion in 2026 on data centers, chips, networking equipment, and related AI infrastructure. This surge targets $818 billion for spending in 2026, according to the analysis from Jefferies. The financial strain is significant; aggregate debt for these companies is projected to rise from $226 billion in 2024 to $502 billion in 2026. Simultaneously, their combined free cash flow is expected to drop from $238 billion in 2024 to a marginal deficit of $2 billion by 2027. Despite these balance sheet pressures, Jefferies maintains a bullish outlook on the AI investment cycle. A key driver is the rapid growth in customer commitments. Backlogs have increased by roughly $1.67 trillion since December 2024, which is nearly four times the increase in capital expenditure during the same period. The firm notes that Anthropic could generate up to $200 billion in annual revenue within seven years of its launch. Jefferies highlights three reasons for its constructive stance, with customer commitments growing faster than investment costs being the first major factor supporting continued infrastructure expansion.
