3 weeks ago
Big Tech to Spend $740 Billion on AI This Year
Four of the biggest technology companies in the world — Amazon, Alphabet, Meta and Microsoft — are planning to spend about $740 billion this year on artificial intelligence, or AI.
That is about three-quarters of a trillion dollars, which is more money than most countries make in a whole year.
AI is like a very smart computer brain, but it needs giant buildings full of computers, called data centres, to work.
Building those data centres costs a lot of money.
Some investors got worried that the companies were spending too much, and the value of some tech companies went down.
Then Microsoft and Amazon showed that their cloud businesses were growing fast, which means more customers are using AI.
The companies are also getting partners like BlackRock and Nvidia to help pay for the projects.
Experts say technology companies used to be cheap to run, but now they are becoming more like factories.
Whether all this spending will pay off is still a big question.
Amazon, Alphabet, Meta and Microsoft are tracking toward roughly $740 billion in combined 2026 AI capital expenditure, up about 77% from $410 billion in 2025.
Amazon leads with about $200 billion, Alphabet guided to $195–205 billion, Meta to $125–145 billion, and Microsoft sits at approximately $120 billion.
In late July, markets turned cautious: Meta shares fell roughly 10% after free cash flow dropped 91% to $784 million, Alphabet's free cash flow went negative at minus $5.9 billion, and over $1.3 trillion was erased from semiconductor company values.
Bullish signals followed: Microsoft's Azure growth accelerated to 43% and Amazon posted second-quarter revenue of $200.6 billion, up 20%, suggesting AI demand is real and growing.
Financing is moving beyond internal funds: Meta shifted a $14 billion data centre to BlackRock-managed funds, Nvidia is reportedly in talks to guarantee about $250 billion for OpenAI's Ohio project, and Uber is committing over $10 billion for 120,000 robotaxis.
- Who
- Amazon, Alphabet, Meta and Microsoft, along with investors and partners such as Nvidia, BlackRock, Uber and OpenAI.
- What
- Combined capital expenditure of roughly $740 billion on AI infrastructure in 2026, up about 77% from $410 billion in 2025.
- Where
- Primarily the United States, including OpenAI's project in Ohio, with Europe also opening a €30 billion AI gigafactory tender.
- When
- 2026, with market sell-offs and earnings reports occurring in late July.
- Why
- The companies are betting that enterprise AI adoption will justify massive infrastructure spending, though investors debate whether the returns will cover the capital consumed.
The Bull Case
The Bear Case
Are AI investments paying off?
The Bull Case
Microsoft's Azure growth accelerated to 43% and Amazon's second-quarter revenue rose 20%, showing enterprise AI adoption is expanding and the demand for AI infrastructure is real.
The Bear Case
Meta's free cash flow fell 91% to $784 million, Alphabet's turned negative at minus $5.9 billion, and over $1.3 trillion was erased from semiconductor company values because investors question whether the spending will earn its cost of capital.
Is the scale of spending justified?
The Bull Case
The roughly $740 billion, up 77% in a year, reflects the industrial scale needed as AI demand grows, with large cloud businesses growing faster than a year earlier.
The Bear Case
Financing is moving beyond what companies can fund internally — Meta offloaded a $14 billion data centre to BlackRock-managed funds and Nvidia may guarantee about $250 billion for OpenAI — signalling the spending may be too large to sustain.
Key facts
- Combined 2026 AI capex
- ~$740 billion
- Year-on-year increase
- ~77% (from $410 billion in 2025)
- Amazon capex
- ~$200 billion
- Alphabet capex guidance
- $195–205 billion
- Meta capex guidance
- $125–145 billion
- Microsoft capex
- ~$120 billion
- Microsoft Azure growth
- 43% (accelerated)
- Amazon Q2 revenue
- $200.6 billion, up 20%









