Amazon boosts CAPEX to $44.2B, up from $25B last year

Amazon just dropped $44.2 billion in a single quarter on capital expenditures. That’s not a typo. The figure represents a 77% increase from the $25 billion the company spent in the same period last year, and it signals that Amazon’s AI spending spree is accelerating, not slowing down.

The numbers behind Amazon’s AI bet

The quarterly figure is eye-catching on its own, but zoom out and the picture gets even more dramatic. Amazon’s trailing twelve-month capex now sits at $147.3 billion, a 67% jump from $88 billion over the prior twelve-month period.

Operating cash flow tells a healthier story, rising 30% to $148.5 billion on a trailing twelve-month basis. Free cash flow, the money left after all that spending, cratered to just $1.2 billion. When your operating cash flow is $148.5 billion and your free cash flow is $1.2 billion, nearly every dollar coming in the door is going right back out.

Why AWS matters more than ever

Amazon Web Services, the company’s cloud computing arm, met revenue expectations for the quarter. AWS is the engine that’s supposed to justify all this spending, and so far, it’s delivering.

Analysts have pointed to AWS’s performance and optimistic forward guidance as the key reason the market isn’t punishing Amazon for burning through nearly all its free cash flow. That reasoning has contributed to a measurable reduction in bearish sentiment around Amazon’s stock.

What this means for investors and the broader market

The investment thesis here is straightforward but carries real risk. Amazon is betting that AI infrastructure will be the foundation of its next decade of growth, primarily through AWS. If enterprise demand for AI compute continues to surge, the $147.3 billion in trailing capex starts looking like a moat, not a money pit.

But the free cash flow compression is hard to ignore. At $1.2 billion on a trailing twelve-month basis, any slowdown in AWS revenue growth, any delay in AI monetization, or any macroeconomic headwind that forces enterprise clients to cut cloud budgets makes the math uncomfortable fast.