Amazon’s $100B AI Strategy Is Bigger Than You Think
Inside the boldest infrastructure play in tech history and how Amazon is rebuilding the rails of the AI economy.
You received this email because you subscribed to Global Data Center Hub, a newsletter about the global data center sector.
Thank you very much for supporting my newsletter this month. Your readership encourages me to provide these insights, and I’m truly grateful for it.
We publish new insights seven days a week, helping you stay ahead of the most important shifts in data centers, AI infrastructure, and global connectivity.
Our premium insights are reserved for paid subscribers.
If you haven’t upgraded yet, now’s a great time:
👉 Subscribe and join hundreds of readers of “Global Data Center Hub” with a 20% discount on the annual plan: Subscribe here
When Amazon released its 2025 shareholder letter, one quote stood out like a thunderclap:
“AI is a once-in-a-lifetime reinvention of everything we know.” Andy Jassy
And with that, Amazon’s CEO defended what may be the most ambitious infrastructure investment in the company’s history:
$100 billion in capital expenditures.
In a single year.
Most of it dedicated to AI.
This isn’t just an experiment.
It’s a transformation.
From Infrastructure to Intelligence
Here’s what Amazon is doing and why it matters:
1,000+ AI applications across retail, streaming, health, and more
A complete rebuild of Alexa, now powered by Anthropic’s Claude
Investment in Trainium2 chips, offering 30–40% better price-performance than GPUs
Triple-digit AI revenue growth, now a multi-billion-dollar run rate
Building AI data centers at scale, faster than anyone expected
In short: Amazon wants to own the infrastructure of the AI economy just like it did with cloud.
Chips, Cloud, and Control
Why Trainium2 matters:
Jassy says the biggest cost in AI today is silicon.
By designing its own chips, Amazon cuts reliance on Nvidia and others and gains pricing leverage at the infrastructure layer.
Jassy didn’t just stop at chips. He outlined a roadmap to lower overall AI costs, including:
Prompt caching
Model distillation
More efficient inference
Better model architectures
All of which reduce the price of deploying AI at scale.
Revenue Today, Moats Tomorrow
It’s not just a long-term bet.
Amazon is already seeing returns:
AI revenue is growing at triple-digit rates
The Claude-powered Alexa+ is rolling out
Enterprise customers are using Amazon’s AI stack in AWS
But the real prize?
Control of the rails.
Amazon is rebuilding its infrastructure so that every product and service, from online shopping to cloud hosting, runs on its own AI stack.
Not OpenAI’s. Not Google’s.
Amazon’s.
Strategic Parallels: The AWS Playbook
If this strategy sounds familiar, it should.
Early AWS investments were viewed as reckless. It took years to generate profits.
But when the demand came, Amazon was ready.
Now, the same logic applies to AI:
The cost curve will drop
Demand will explode
The company with infrastructure in place will win
This is Jassy’s real bet:
That infrastructure, not just innovation, defines leadership.
Final Thought
Amazon isn’t just investing in AI.
It’s preparing to control it.
The company is making the largest annual AI infrastructure investment in history—because it sees what’s coming next:
Not just smarter models.
Not just better assistants.
A world where every customer interaction runs on AI.
And a future where whoever owns the infrastructure owns the future.
One More Thing
I publish daily on data center investing, AI infrastructure, and the trends reshaping global data center markets.
Join 900+ investors, operators, and innovators getting fresh insights every day and upgrade anytime to unlock premium research trusted by leading investors and developers.