Amazon 2026

Amazon is significantly ramping up CAPEX. The company is guiding for around $200B in capital expenditures in 2026, up more than 50% YoY (vs. about $131B in 2025), well above market expectations. Most of this spend goes into infrastructure scale-up: AWS data centers, AI compute capacity, custom silicon, automation/robotics, and satellite projects. In short — heavy investment in physical infrastructure and chips to expand compute power and control more of the tech stack. The market reacted negatively, with the stock dropping as much as 9–11% after the announcement. Short term, this puts pressure on free cash flow and margins. Long term, it’s a clear bet on cloud and AI dominance.

AWS over the next four years will be primarily about scale, AI, and infrastructure depth. First, AWS will continue expanding data center capacity globally. Demand for AI training and inference compute is structurally increasing, and enterprises are migrating more critical workloads to the cloud. More regions, more availability zones, more raw compute. Second, custom silicon will matter more. AWS is investing heavily in its own chips (e.g., for AI acceleration and general compute) to reduce dependency on third-party suppliers and improve margins. Vertical integration becomes a competitive advantage. Third, enterprise AI adoption will move from experimentation to production. That means higher, more stable workloads: model hosting, data pipelines, fine-tuning, and large-scale inference. This should support durable revenue growth. Short term, margins may face pressure due to aggressive CAPEX. Long term (3–4 years), if utilization rates stay high, operating leverage could significantly improve.

Amazon Web Services currently generates around $100B in annual revenue, grows roughly 12–20% year over year, operates at 25–35% margins, and could reach $160–190B in revenue by 2028 if it sustains a 15–20% CAGR, while Amazon’s total CAPEX of $130–200B per year heavily supports AWS infrastructure possible Cloude expansion.