Amazon's third-quarter 2026 earnings delivered a number that very few analysts expected to see confirmed in hard numbers: AWS (Amazon Web Services) has reached an annualized revenue run rate of $184 billion, growing 39.3% year-over-year. It is the fastest acceleration in nearly five years and reinforces the thesis that cloud computing is not slowing down — it is expanding at a pace that financial markets are still trying to digest.
Andy Jassy, Amazon's CEO and the founder of the very cloud division that now generates the bulk of the company's profit, made no secrets during the earnings call. "There has never been a $1 trillion revenue business," he said, "and we may have two of them in the next 5-10 years." The explicit reference to "two" suggests he sees the same potential in another segment — likely Amazon.com (global ecommerce), whose operating margin grew alongside AWS, or perhaps Prime Video, which the company itself says now operates at a "tens of billions" revenue run rate and is "now profitable in its own right."
But the path to $1 trillion starts with $220 billion in capex for 2026. Amazon has already spent $75.9 billion in the third quarter alone, most invested in AWS data center expansion. The company signaled it will spend "meaningfully more" in 2027. To put that number in perspective: Amazon spends more on IT infrastructure than many countries spend on education. The memory investment, which raised the capex projection from $200 to $220 billion, is a concrete indicator of how intense the race for AI processing capacity has become.
The math behind this investment is what AWS executives call the "flywheel effect." Matt Wood, AWS's AI leader, explained that "the more customers use AI, the more they use core AWS services" — more agents drive more CPUs, more CPUs drive more data, more data feeds more AI models. It is a virtuous cycle the company believes is in the "very early days" of enterprise adoption. Jassy countered similar skepticism about cloud migration by noting that "about 85% of global IT spend is still on-premises" and that we are in a "10- to 20-year cycle" where that equation will flip.
What is interesting to observe is the composition of revenue. AWS is not growing simply because enterprises are migrating traditional workloads — Bedrock, the multi-model AI service, is already the "fastest-growing multi-billion dollar service in AWS history." Lambda, the serverless platform, just topped $1 billion in annualized revenue. And AgentCore recorded in Q3 twice as many agent deployments as the prior three quarters combined. In other words: AI is not replacing traditional cloud, it is multiplying demand for it.
The market, however, did not react the way AWS investors hoped. Amazon shares fell about 6% in after-hours trading, indicating that the cost of financing this expansion is being viewed as a risk — not necessarily because the growth is not happening, but because the timing of returns is uncertain. Dan Salmon from New Street Research noted that "despite AWS revenue and operating income beating consensus estimates, overall revenue fell short of buyside expectations." In other words: the market is discounting the uncertainty that $220 billion in capex will translate into enough revenue to cover the investment.
To compete, AWS signed new agreements with Meta, NVIDIA, OpenAI, Salesforce, Oracle, Qualcomm, Verizon, ADP, and Stripe during the quarter. The number of partners is impressive and reveals something important: AWS is not competing for cloud infrastructure, it is competing to become the platform where enterprise AI happens. When you subscribe to Bedrock, when you run agents on AgentCore, when you scale Lambda to serve global applications — you are not just choosing storage or compute, you are choosing an ecosystem.
The question is whether the cloud market is ready to sustain this $1 trillion bet? Synergy Research Group estimates that quarterly infrastructure services revenue has already surpassed $45.4 billion, with a trailing twelve-month total of $164 billion. Cloud services growth hit 55% in Q3, driven mainly by AI. If AWS can maintain 26% market share (versus Microsoft's 19% and Google's 16%), and if each emerging competitor (Oracle, Alibaba, xAI, CoreWeave) stays between 2% and 4%, the math seems possible — but execution is what separates promises from trillions in real revenue.
Sources: TechCrunch, The Register, Fierce Network
✓ Independent sources cross-checked and verified before publishing
