Big tech companies are spending $725 billion on AI infrastructure in 2026, triggering growing AI bubble fears.
Amazon, Google, Microsoft, and Meta account for the vast majority of that combined capital expenditure.
Per Bloomberg, spending has surpassed $700 billion and continues accelerating into the second half of 2026.
Who Is Spending the Most on AI Infrastructure
Amazon leads with projected 2026 capex of $200 billion, up from $131 billion spent in 2025.
Google is close behind at $175-185 billion, more than doubling its 2025 investment of $91 billion.
Microsoft and Meta follow with capital intensity ratios above 45% of their total annual revenues.
Oracle rounds out the five-company group, making the combined AI infrastructure spend roughly $660-690 billion.
SoftBank separately plans a $100 billion Roze AI robotics IPO targeting the second half of 2026.
Is the $725 Billion AI Spending a Bubble?
J.P. Morgan estimates the industry needs $650 billion in new AI revenue annually just to earn a 10% return.
Current AI-linked revenues run between $50 and $150 billion per year, a 4x to 13x shortfall.
The Federal Reserve named AI as a top systemic financial risk in 2026, just behind geopolitical threats.
Capital intensity ratios above 50% at Oracle and Meta would have been unthinkable in any prior decade.
Despite the investment, AI has not yet produced a measurable impact on U.S. GDP growth as of mid-2026.
What the Revenue Picture Actually Looks Like
Microsoft reported its AI business is running at a $37 billion annual revenue rate, up 123% year-over-year.
Alphabet’s cloud revenue surged over 60%, and its cloud backlog nearly doubled to $460 billion last quarter.
These numbers are strong but remain a fraction of the $725 billion being invested in new infrastructure.
Critics say if revenue growth slows, companies will face painful write-downs on overbuilt data center capacity.
Supporters counter that AI’s compounding returns will justify the upfront spending over the next five years.
How This Spending Wave Is Reshaping the Tech Industry
Despite the massive capex, tech layoffs hit 153,000 workers in 2026 as firms cut headcount to fund AI.
NVIDIA is the biggest beneficiary, supplying the H200 and Blackwell GPUs powering most new data centers.
Electricity demand from AI data centers is straining power grids across the U.S., Europe, and Southeast Asia.
The capex race is directly driving agentic AI development as firms race to monetize their infrastructure.
Data center construction, power infrastructure, and semiconductor supply chains are running at maximum capacity.
What Analysts Predict for AI Spending in 2027 and Beyond
CNBC reported combined big tech capital expenditure is forecast to top $1 trillion by 2027.
Spending is expected to slow only if a clear AI model capability plateau becomes apparent to investors.
No such plateau signal has appeared yet, and all four major hyperscalers have raised 2026 spending guidance.
The consensus view is that AI infrastructure spending will remain elevated through at least the end of 2028.
The critical question is whether enterprise AI adoption revenue catches up with infrastructure investment in time.
Related Articles
Trust Post Desk
A journalist and editor at TrustPost.org covering world and national news, technology updates and human-interest stories. They check every fact, interview sources in person or online, and aim to deliver clear, accurate reporting. Their work ranges from breaking news to in-depth features and daily newsletters. Outside the newsroom, they follow emerging trends and engage with readers on social media.
