AI Capex Surge Expected to Continue as Hyperscalers Spend $700B in 2026
Event summary
- U.S. hyperscalers spent an estimated $400 billion on AI-related capital expenditures in 2025, with expectations to reach $700 billion in 2026.
- CoBank's report suggests AI infrastructure spending has a long runway due to strong returns on invested capital and robust supply chain guidance.
- NVIDIA projected more than $1 trillion in cumulative revenue for its AI GPU chipset business through 2027, up from approximately $130 billion in 2025.
- Micron reported Q2 revenue of $23 billion, a 196% year-over-year increase, driven by AI infrastructure spending.
- Anthropic's revenue run rate is expected to reach $50 billion by the end of June 2026, up from $9 billion at the end of 2025.
The big picture
The AI capex cycle shows no signs of slowing down, with hyperscalers continuing to invest heavily in next-generation digital technologies. The interconnected nature of the AI ecosystem and the circular relationships between participants present risks, but current indicators suggest these risks are isolated rather than systemic. The transition from foundational infrastructure to AI inference applications is expected to drive further growth and innovation in the sector.
What we're watching
- Market Concentration
- How the highly concentrated nature of AI spending will impact market stability and potential risks.
- Profitability Sustainability
- Whether current profitability levels across the AI ecosystem can be maintained amid rapid growth.
- Economic Impact
- The pace at which AI-driven productivity gains will translate into broader economic growth.
Related topics