AnswersPublished24 sources
Why Andy Jassy Thinks AWS Could Become a $1 Trillion Business
Jassy says AWS could eventually reach a $1 trillion annual revenue run rate, but that is a long term possibility, not a dated forecast. AWS brought in $42.2 billion in second quarter 2026 revenue, up 37% year over year, while Amazon reported a $496 billion backlog.
Published byEdited with GPT-6 LunaImages generated with GPT Image 2
Research answer
Key takeaways
- Jassy says AWS could eventually reach a $1 trillion annual revenue run rate, but that is a long term possibility, not a dated forecast.
- AWS brought in $42.2 billion in second quarter 2026 revenue, up 37% year over year, while Amazon reported a $496 billion backlog.
- Jassy argues that AI inference costs need to fall roughly tenfold to make AI practical across more business applications; Amazon sees its chips as one way to help lower those costs.[5][19]
AI Prompt
openai.comCreate a landscape editorial hero image for this Studio Global article: What did Amazon CEO Andy Jassy tell Ben Horowitz and Erik Torenberg in the October 23, 2026 episode of The a16z Show about AWS potentially r. Article summary: Jassy’s message was that AI demand could eventually help make AWS a $1 trillion annual-revenue-run-rate business, but even Amazon’s roughly $220 billion infrastructure investment in 2026 is not enough to meet current dem. Topic tags: general, news, general web, user generated. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts w
Andy Jassy’s $1 trillion vision for AWS is a long-term possibility, not a dated revenue forecast. The near-term challenge is more concrete: AI demand is running ahead of available cloud capacity, even as Amazon plans about $220 billion in capital spending for 2026.1
3
11
The trillion-dollar figure is a possibility, not a timeline
After Amazon’s second-quarter results, Jassy said AWS could “very possibly” become a trillion-dollar annual revenue business “in time.”11 He made a similar case in an October 23 interview with Ben Horowitz and Erik Torenberg: the AI infrastructure buildout could ultimately help AWS reach a $1 trillion annual run rate.
3
That is an expression of potential, not a forecast for a particular year. AWS reported $42.2 billion in revenue for the second quarter of 2026, up 37% year over year. That was above the $40.54 billion StreetAccount estimate cited by CNBC.2
12 The reported growth rate was also up from 28% in the first quarter.
27
32
These figures show a rapidly growing business and substantial customer commitments. They do not, by themselves, establish when AWS might reach $1 trillion or how much of its backlog will turn into revenue in any particular year.
Why $220 billion still does not solve the capacity shortage
On October 23, Jassy said AWS capacity was largely sold out through 2027, with significant customer commitments extending into 2028.1 In July, he had also said Amazon would still lack enough computing capacity to serve all the demand it expected in 2026, despite raising its full-year capital-spending plan.
2
Amazon increased its planned 2026 cash capital expenditure from about $200 billion to roughly $220 billion. Jassy attributed the higher estimate in part to rising memory costs.11 But a larger spending plan is not the same as capacity already online: investment has to become usable infrastructure before it can serve customers. The company’s comments point to a mismatch between the pace of demand and the time required to add capacity.
1
2
The backlog and multi-year commitments give Amazon visibility into demand, but they do not mean every project is already generating revenue. The available reporting does not establish the precise timing of revenue from those commitments.1
2
Why inference costs matter to the growth case
Inference is the process of running a trained AI model to produce results, and its cost affects whether businesses can use AI broadly. Jassy has argued that inference remains roughly an order of magnitude too expensive for AI to be used as widely as it could be in business software.19
The commercial logic is straightforward: if inference gets cheaper, more applications may become economical for businesses. Jassy has linked Amazon’s custom chips to lowering those costs and making broader adoption more viable.5
19 That is a potential source of future demand for AWS—not proof that lower costs or wider adoption are guaranteed.
The trade-off: growth now, cash returns later
The infrastructure buildout comes with a near-term cash cost. Amazon’s trailing-12-month free cash flow was reported at negative $7.6 billion, with coverage attributing the pressure to spending on data centers and other infrastructure.40
45 The underlying timing risk is that capital outlays arrive before new facilities can support additional customer usage and revenue. Reporting has noted that free-cash-flow pressure may persist until new data centers come online.
41
That makes the central question more than whether customers want AI capacity: it is whether Amazon can bring enough capacity online, turn commitments into sales, and eventually generate cash returns that justify the upfront investment. Strong AWS growth and a large backlog support the demand side of the case; the spending and cash-flow figures show the execution challenge.2
27
41
What the evidence does—and does not—show
One summary of the October interview describes Jassy discussing faster product development, including a Bedrock team rebuilding an inference engine with a small engineering group.17 That example may illustrate how Amazon hopes to move faster internally, but it does not change the distinction between software development speed and the time needed to build physical cloud capacity.
The sources available here do not verify specific claims about a reworked Alexa, nor do they show what share of the 2026 spending plan is tied to commitments expected to generate revenue in 2027 or 2028. Those details should not be treated as established evidence for the trillion-dollar outlook.
The supported case is narrower: AWS is growing quickly, customer demand is stretching beyond available capacity, and Amazon is investing heavily to expand supply. Jassy’s $1 trillion ambition depends on that buildout—and on making AI use economical enough to sustain broader demand.1
2
3
19
Studio Global AI
Continue your research
This page includes a source-backed answer you can continue inside Studio Global.
People also ask
What is the short answer to "Why Andy Jassy Thinks AWS Could Become a $1 Trillion Business"?
Jassy says AWS could eventually reach a $1 trillion annual revenue run rate, but that is a long term possibility, not a dated forecast.
What are the key points to validate first?
Jassy says AWS could eventually reach a $1 trillion annual revenue run rate, but that is a long term possibility, not a dated forecast. AWS brought in $42.2 billion in second quarter 2026 revenue, up 37% year over year, while Amazon reported a $496 billion backlog.
What should I do next in practice?
Jassy argues that AI inference costs need to fall roughly tenfold to make AI practical across more business applications; Amazon sees its chips as one way to help lower those costs.[5][19]