Amazon Web Services gave Amazon its clearest lift in years in the first quarter of 2026, with revenue rising 28% from a year earlier to $37.6 billion, the fastest growth in 15 quarters. That surge came as Amazon poured $44.2 billion into capital projects, a pace that underscores how aggressively it is building for AI.
The move matters now because Amazon has been one of the megacap laggards of 2026, with its stock about 12% below its 52-week high even as the core cloud business accelerated. Investors searching for signs that the spending is translating into momentum found them in AWS, where operating income reached $14.2 billion on a 37.7% margin, and in Total Amazon, where revenue rose 17% to $181.5 billion and operating income jumped to $23.9 billion.
That scale explains why the quarter drew attention beyond the headline numbers. AWS is Amazon's most important profit engine, and the latest results suggest companies are increasingly training and running AI models where their data already sits. For many of them, that means AWS. Amazon's own spending is aimed mostly at AI infrastructure, and the company is leaning on the same habit that has defined earlier turns in its history: making large bets before the payoff is obvious.
The friction is that the cash story has moved in the opposite direction. Amazon's free cash flow fell to about $1.2 billion over the trailing 12 months from nearly $26 billion before the decline, even as the business is said to be in its best shape in years. The math is blunt: the $44.2 billion Amazon spent on capital projects in the first quarter was nearly double the $25 billion it spent a year earlier, a jump of about $19.2 billion, or roughly 77%.
Amazon's broader business still looked solid. North America revenue rose 12% to $104 billion, the international business grew 19%, and custom silicon now runs at more than a $20 billion annual revenue pace while growing at triple-digit rates. But the open question is whether Amazon keeps pushing capital spending at this pace. For shareholders, that is the issue that now matters most: whether AWS can keep reaccelerating fast enough to justify the cash drain, or whether the AI buildout will keep squeezing free cash flow before it turns into a larger payoff.
