AI Capex Supercycle Accelerates in 2026 as Big Tech Commits $725 Billion to Data Centers, Up 77% in a Single Year

The four largest US hyperscalers will spend roughly $725 billion on capital expenditures in 2026, up about 77% from $410 billion in 2025, according to Bloomberg and company guidance.

Amazon, Microsoft, Alphabet and Meta have all raised or confirmed massive AI infrastructure budgets, making this the largest concentrated corporate investment cycle in modern history.

The spending surge is driven by insatiable demand for cloud and AI computing capacity. Cloud revenue growth accelerated across all three major providers in the second quarter, with Google Cloud up 82%, Microsoft Azure up 43% and AWS up 37%.

Key Takeaways: Big Tech AI Capex in 2026

• Combined 2026 capex: roughly $725 billion across Amazon, Microsoft, Alphabet and Meta, up 77% from $410 billion in 2025.
• Amazon: raised guidance to $220 billion from $200 billion, citing AI infrastructure and memory costs.
• Microsoft: guided to $190 billion for 2026, a 61% increase from the previous year.
• Alphabet: raised 2026 capex to $195–205 billion from $180–190 billion.
• Meta: 2026 capex guidance of $125–145 billion, roughly double its 2025 spend.
• Cloud growth: Google Cloud +82%, Azure +43%, AWS +37% year over year in Q2 2026.
• Backlog: remaining performance obligations across major cloud providers near $1.7 trillion.
• Free cash flow pressure: Amazon Q1 2026 FCF fell 95% to $1.2 billion as capex reached $44.2 billion.

Company-by-Company: Who Is Spending What in 2026?

The table below compares 2026 capex guidance across the four largest hyperscalers, along with their 2025 spending and year-over-year growth.

UBS estimates total hyperscaler capex will reach $673 billion in 2026, rising to $843 billion in 2027 and $894 billion in 2028, though growth slows sharply from 76% in 2026 to 25% in 2027 and 6% in 2028.

Why Is Big Tech Spending So Much on AI Infrastructure?

Three forces are driving the spending supercycle, and each one reinforces the others.

1. Cloud demand is accelerating, not slowing

Cloud revenue growth reaccelerated across all major providers in 2026. Google Cloud grew 82% year over year, its fastest rate among the three, while AWS posted its fastest growth in 18 quarters at 37%. These growth rates signal that enterprise AI adoption is still in early innings.

2. Capacity remains the binding constraint

Amazon CEO Andy Jassy said that even at $220 billion in spending, the company will not have enough capacity to meet demand, an imbalance that could persist into 2027. Microsoft similarly indicated that AI capacity is expected to remain constrained through 2026.

3. Vertical integration is becoming a competitive advantage

Each hyperscaler is building integrated AI stacks. Alphabet combines proprietary TPUs and CPUs with Gemini and Google Cloud. Microsoft integrates cloud infrastructure with both proprietary and third-party AI models. Amazon pairs AWS with Trainium, Inferentia and Graviton chips. These integrated systems allow better performance and cost optimization.

How Does the AI Capex Surge Affect Free Cash Flow and Investors?

The spending surge is straining free cash flow across the board. Amazon reported first-quarter free cash flow of just $1.2 billion, down roughly 95% year over year, as quarterly capex reached $44.2 billion. Meta generated $12.4 billion of free cash flow in the same quarter against $19 billion of capex, and some models project its free cash flow turning negative by 2027.

Capex as a percentage of sales has reached extreme levels at some companies. Oracle leads at 86% of sales, followed by Meta at 54%, Microsoft at 47%, Alphabet at 46% and Amazon at 25%.

For investors, the key question is return on invested capital. The market has begun to reward both chipmakers and hyperscalers simultaneously, suggesting confidence that the ecosystem itself is becoming more valuable. Remaining performance obligations near $1.7 trillion provide visibility into future revenue, and as older contracts reprice higher, AI investment can increasingly be funded by AI-generated cash flow rather than new debt.

What Does This Mean for the Broader Economy?

The AI capex boom has become a significant contributor to US economic growth. Private data center construction spending rose from $15.59 billion in January 2023 to $59.31 billion in May 2026 on a seasonally adjusted annualized basis, roughly 3.8 times higher, according to Federal Reserve data.

Beyond direct construction, the spending ripples through semiconductor supply chains, power generation, networking equipment and cooling infrastructure. A single modern 100-megawatt AI data center can cost more than $4 billion, including chips.

The scale of investment also raises questions about depreciation and returns. Roughly half of the 2026 spend funds servers and silicon that depreciate over five to six years, while the rest funds data center shells and power infrastructure with longer useful lives. A depreciation wall lands between 2027 and 2029, and how the market absorbs it will shape tech earnings for years.

Will AI Capex Keep Growing After 2026?

J.P. Morgan estimates hyperscaler capital expenditures will exceed $1 trillion in 2027 and beyond, four times the level seen before the AI boom. PwC projects global data center spending could reach $31.6 trillion through 2050, with annual spending rising from roughly $800 billion in 2026 to $1.1 trillion in 2030 and $1.8 trillion by 2050.

However, growth rates are expected to decelerate. UBS forecasts capex growth slowing to 25% in 2027 and 6% in 2028, after 76% in 2026. The transition from hypergrowth to sustained high-level spending will test whether AI revenue can justify the balance sheet commitments already made.

Conclusion: The Largest Corporate Investment Cycle in Modern History

2026 marks the year Big Tech's AI spending moved from ambitious to unprecedented. At $725 billion combined, the four largest hyperscalers are spending more on AI infrastructure than the GDP of most countries.

The bet is straightforward: build the capacity now, capture the cloud and AI market for the next decade. Whether that bet pays off depends on adoption, monetization and the durability of demand. For now, the hyperscalers themselves say the biggest risk is not spending too much, but spending too little.

Frequently Asked Questions (FAQ)

How much are Big Tech companies spending on AI in 2026?

Amazon, Microsoft, Alphabet and Meta are collectively spending roughly $725 billion on capital expenditures in 2026, up about 77% from $410 billion in 2025. Amazon leads with $220 billion, followed by Alphabet at $195–205 billion, Microsoft at $190 billion and Meta at $125–145 billion.

Why is AI capex rising so fast in 2026?

AI capex is rising because cloud and AI demand is accelerating faster than capacity can be built. Google Cloud revenue grew 82% year over year, AWS grew 37% and Azure grew 43% in Q2 2026. Amazon CEO Andy Jassy said the company will not have enough capacity to meet demand even at $220 billion in spending.

Does the AI capex boom hurt free cash flow at Big Tech companies?

Yes, free cash flow is under significant pressure. Amazon's Q1 2026 free cash flow fell 95% to $1.2 billion as capex reached $44.2 billion. Meta generated $12.4 billion of free cash flow against $19 billion of capex, and some analysts project its free cash flow turning negative by 2027.

What does the AI capex surge mean for investors?

For investors, the key metric is return on invested capital. While capex has surged, remaining performance obligations near $1.7 trillion provide revenue visibility, and cloud growth is accelerating. Markets have begun rewarding both chipmakers and hyperscalers, suggesting confidence in the ecosystem. However, a depreciation wall between 2027 and 2029 will test earnings.

Will AI capex continue to grow after 2026?

J.P. Morgan estimates hyperscaler capex will exceed $1 trillion in 2027 and beyond, and PwC projects global data center spending could reach $31.6 trillion through 2050. However, UBS forecasts growth slowing to 25% in 2027 and 6% in 2028, after 76% in 2026, as the cycle matures.