Amazon Among Big Tech Firms Spending $660 Billion On AI Infrastructure
Big Tech, including Amazon, is allocating roughly $660 billion in 2026 to build AI data centers and infrastructure, shifting from software to capital-intensive builds. Several firms, Amazon among them, are approaching or exceeding operating cash flow limits, forcing billions in bond borrowing and risking infrastructure overcapacity.
1. Record AI Infrastructure Commitments
Big Tech companies plan to invest a combined $660 billion in AI infrastructure in 2026, marking a major pivot from asset‐light software to physical data center and power plant builds. Amazon Web Services is a key driver of this spending, aiming to scale high-performance compute capacity for machine learning workloads.
2. Amazon's Financing Strategy
With AI capital requirements surging, Amazon is nearing its operating cash flow limits and has begun issuing multi-year bonds to fund data center expansion. These bond issuances are intended to bridge shortfalls between cash flow and upfront capex needs for servers, networking gear and specialized AI chips.
3. Overcapacity And Return Risks
Unlike software rollouts, data centers represent long-term, irreversible commitments that may lead to excess capacity if demand projections fall short. Historical infrastructure booms suggest potential for cost overruns, underutilized assets and write-downs if AI adoption does not scale as expected.