Can Amazon’s big AI bet deliver for investors?

For many people Amazon is a ubiquitous part of their day-to-day lives. The familiar brown cardboard boxes with their signature black smile logo making their way to doorsteps up and down the country at regular intervals.

 

But while Amazon is principally known at a consumer level for its e-commerce operations, it is the much higher margin Amazon Web Services (AWS) business which is the key engine of growth, and it is here where much of the excitement around the stock can be found. While AWS represents less than 20% of sales, it accounts for nearly 60% of its profit.

A major player in AI

Amazon is a major participant in the AI arms race, one of the so-called ‘hyperscalers’ spending hundreds of billions of dollars on the infrastructure behind the artificial intelligence push.

The key question for investors is whether that spending will deliver commensurate returns or if, instead, Amazon ends up being unable to justify the scale of the outlays.

The market seems to have been at least partly won over by recent earnings reports which hint that Amazon is starting to see some tangible benefits from this mountain of capital expenditure.

The improved sentiment is reflected in its performance in 2026 to date, where it has outpaced other hyperscalers like Alphabet, Meta and Microsoft. Even if it is still lagging these rivals on a five-year view. Both on a total returns basis.

 

The company has progressed the development of its own in-house Trainium chips and is also expanding its Bedrock cloud-based service which allows clients to use powerful generative AI models without building or managing their own heavy computer hardware.

Lowering the costs of artificial intelligence

In the 2025 annual report, CEO Andy Jassy commented: “Virtually all AI thus far has been done on Nvidia chips, but a new shift has started. We have a strong partnership with Nvidia, will always have customers who choose to run Nvidia, and we will continue to make AWS the best place to run Nvidia. However, customers want better price-performance.

“Having our own hotly demanded AI chip opens up many possibilities, but perhaps none larger than the ability to lower costs for customers and secure better economics for AWS. At scale, we expect Trainium will save us tens of billions of capex dollars per year, and provide several hundred basis points of operating margin advantage versus relying on others’ chips for inference.”

The idea that Amazon might make access to AI more affordable and accessible is an interesting one which connects with the value credentials which have arguably been one of the key drivers for its e-commerce arm. Although a risk for the business is the competitive threat posed by Chinese AI operators which have demonstrated their own ability to reduce the costs of the technology.


What is agentic AI?

A system that makes its own decisions and takes independent actions to reach a specified goal without significant human input.


 

Backing for the strategy

Bank of America analyst Justin Post thinks Amazon’s long-term strategy can work: “We believe Amazon’s AI positioning has improved significantly in the past 12 months, with AI revenues growing to 15% of total Cloud revenues, benefiting from rapid capacity additions, Trainium improvements, and accelerating Bedrock demand.

“While higher 2026 capex will pressure free cash flow, we believe Amazon is building an asset base that will have high returns, and that Amazon is well positioned to further benefit from the upcoming inference wave driven by agentic AI. We expect multiple years of 20% growth ahead for AWS, driven by multi-year corporate adoption of AI.”

Discussing the latest quarterly numbers from July, BofA’s Post notes that while Amazon increased its capital expenditure projection for 2026 by $20 billion to $220 billion – partly thanks to the industry-wide increase in memory chip costs – it also provided a clear indication of the returns on investment from data centres.

Buildings are expected to have a 30-year life, servers and networking assets are expected to have payback periods of less than three years and strong returns over the remaining three years of their useful lives. Post also highlights an encouraging increase in AWS margins from 32.9% and 39%.

 

What about other areas of the business?

Outside of AI, Amazon is seeing strong advertising growth – with ad sales of $19.8 billion in the second quarter. These come from expanded advertising on its Prime Video streaming platform and sponsored products on Amazon Prime. The e-commerce business as a whole continued to deliver strong growth as the company benefits from offering an increased volume of logistics services to third parties.

Comparing its valuation to other hyperscalers, Amazon trades at a discount to Microsoft but a premium to Meta and Alphabet as the chart shows. Though its forecast price to earnings ratio is at a two-year low.

Tom Sieber

Tom Sieber: Content Editor

Tom Sieber is AJ Bell's Content Editor. He was previously the Editor of Shares Magazine. He has been with the business since 2012.

Tom is a regular contributor to the AJ Bell Money & Markets...

These articles are for information purposes and should only be used as part of your investment research. They aren't offering financial advice and past performance is not a guide to future performance, so please make sure you're comfortable with the risks before investing.