Microsoft, Google and Amazon Dominate AI with Billions, Infrastructure and Energy

Thursday, 30 April 2026 at 12:00

Microsoft, Google en Amazon domineren AI met miljarden, infrastructuur en energie

Microsoft owns roughly 27 percent of OpenAI and Google about 14 percent of Anthropic, while Amazon is estimated to hold 15 to 19 percent of Anthropic. These stakes show the AI race isn’t just about better models—it’s about controlling infrastructure, energy, and compute.

The numbers reveal a deeper battle behind products like ChatGPT, Claude, and Gemini. Tech giants aren’t just investors; they’re owners of the capacity that makes AI possible.

What do these ownership stakes really mean?

The AI race is about controlling both technology and distribution. Big Tech is deliberately investing in ecosystems where they supply the foundations themselves.

OpenAI’s split looks like this:

Microsoft: roughly 27 percent
OpenAI Foundation: roughly 26 percent
Employees and other investors: roughly 47 percent

For Anthropic:

Google: roughly 14 percent
Amazon: estimated 15 to 19 percent
Founders, employees, venture capital, and other investors: the remaining stake

These structures show that influence doesn’t only come from a majority stake, but from a strategic position inside the network.

Why are tech giants pouring in billions?

The key reason is control of infrastructure. Companies like Microsoft, Google, and Amazon provide not just cash, but access to their cloud platforms.

Here’s how that works:

AI models run on massive compute in data centers
Investments often include cloud credits and infrastructure
AI companies become dependent on a single platform

As widely noted in debates around these deals, the real value often isn’t cash—it’s discounted access to data center capacity.

Why is AI also about energy and compute?

AI development fundamentally hinges on energy and compute. Modern models demand huge amounts of electricity and specialized chips.

In practice, that means:

Training large models costs millions in energy alone
Data centers run 24/7 and keep drawing more power
Access to chips and grid capacity determines who can scale

Microsoft, Google, and Amazon rank among the world’s largest data center operators. They control not just software, but the physical infrastructure and energy supply behind AI.

The AI race is therefore an energy race: whoever can deploy the most compute and power builds the strongest systems.

How is this reshaping AI competition?

The market is shifting from model competition to system competition. Products appear to compete on features, but in reality ecosystems are going head-to-head.

That has three immediate effects:

Compute drives innovation Companies with more compute can train new models faster.
Energy becomes strategic Access to cheap, reliable power shapes AI’s cost base.
Cloud lock-in deepens AI companies stay tied to their investors’ infrastructure.

Has the AI race become mostly financial?

The AI race has morphed into a battle over capital and infrastructure. Investment is no longer just about software, but about physical capacity.

A potential multi‑tens‑of‑billions investment by Google in Anthropic underscores the point. It’s not just about improving Claude—it’s about access to and control over the systems underneath.

Bottom line: AI is about power, hardware, and megawatts

The visible fight between ChatGPT, Claude, and Gemini is only the surface. Beneath it lies a complex power structure of ownership, cloud platforms, energy, and compute.

AI isn’t just a technological shift—it’s a contest over who owns the infrastructure that powers the future.