Jassy Sees a Trillion-Dollar AWS as Amazon Lifts Capex to $220 Billion
Amazon CEO Andy Jassy said on October 29, 2026 that Amazon Web Services can become a trillion-dollar revenue business within the next five to 10 years. The same day, Amazon said it expects 2026 cash capital expenditures (capex, the money spent on physical assets such as data centers, servers and chips) of approximately $220 billion, with the majority directed to AWS. Amazon shares were down about 6% in after-market trading, the extended session after the close, following the earnings call, which was scheduled for 2:00 PM PT.
The AWS figures behind the claim come from the third-quarter release. AWS segment sales reached $46.0 billion, up 39% year over year. On the call, Amazon put the business at a $184 billion annualized revenue run-rate, which is the latest quarter's revenue multiplied by four. The growth rate has moved up from the prior quarter, when AWS net sales rose 37%, according to Amazon's second-quarter announcement.
Jassy also forecast that AWS will grow north of 40% in each of 2027 and 2028. For context, AWS segment operating income was $14.2 billion in the first quarter of 2026, compared with $11.5 billion in the same quarter of 2025. Amazon's first-quarter net sales were $181.5 billion, up 17%.
The capex figure is the larger change. In its second-quarter update, Amazon had guided to roughly $200 billion in 2026 cash capex. It has now raised that estimate by $20 billion to approximately $220 billion. Amazon attributed the increase in part to the higher cost of memory, the chips that hold data while servers run. Amazon also expects to spend meaningfully more on capex in 2027 than in 2026.
Demand is the stated reason for the build-out. In a Reuters report from October 23, Jassy said Amazon's cloud capacity is largely sold out through 2027, with 2026 capacity fully sold and most major 2027 capacity commitments booked. Earlier in the year, Amazon announced a $1 billion investment to create AWS Forward Deployed alongside its second-quarter results, when it reported $200.6 billion in total revenue.
In its third-quarter release, Amazon included a Jassy statement that there has never been a trillion-dollar revenue business and that Amazon may have two of them within five to 10 years. That figure is a company projection, not a segment forecast.
In my view, the arithmetic is the useful place to start. Compounding the $184 billion run-rate at 40% a year crosses $1 trillion in roughly five years. At 20% a year, it takes closer to ten. Jassy's stated forecast covers only 2027 and 2028, so the outer years of the claim depend on growth rates Amazon has not put on the record. The capex line adds a second variable. Building capacity ahead of demand works a little like constructing a toll road before the traffic arrives: the cost is paid up front, and the return depends on the traffic showing up on schedule. Spending of roughly $220 billion in one year is large relative to a segment that currently runs at $184 billion annualized.
The broader context here is a cloud market in which capacity, rather than customer interest, has become the constraint. If Amazon's capacity commitments hold, the next two years of AWS growth will be visible in backlog (contracted orders not yet recognized as revenue) and in the timing of new data center and accelerator deployments. Accelerators are specialized chips used for AI workloads. The market reaction on October 29 suggests investors are weighing that same tradeoff, with the spending forecast and the growth forecast arriving together.
For operators reading this, the practical question is less whether a trillion-dollar AWS is plausible on a five-to-ten-year horizon and more what the next two years of spending buys. The answer will show up in the quarterly results that follow.
