🔍 Read the full analysis: AI Data Centers Are Expanding. Where Is The Money Coming From? on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment in the third quarter of 2026, more than twice the total a year earlier, according to figures compiled from company filings. Their financing mix includes operating cash, debt, leases, partnerships and equity, while the scale of future lease payments and power commitments raises questions about who bears the risk if AI revenue falls short.
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment in the third quarter of 2026, more than twice their combined spending a year earlier, according to calculations based on their filings. The companies are financing the expanding data center build-out with a mix of cash, borrowing, leases, partnerships and share sales, spreading some of the cost beyond their own balance sheets, as discussed in four hard questions facing U.S. data centers.
The companies’ reported capital spending shows the scale of the increase. Amazon spent $77.3 billion on property and equipment from July through September, compared with $35.1 billion in the same period a year earlier. Alphabet reported $51.4 billion, Microsoft $39.4 billion in cash spending, and Meta $33.9 billion. Microsoft’s total rises to $45.2 billion when new finance leases are included. The combined $202 billion figure is the source author’s sum of company cash-flow statements; it is not a single figure reported by the companies.
Company guidance points to continued heavy outlays amid the approaching power bottleneck for AI data centers. Amazon said it expects about $220 billion in cash capital spending for 2026 and that the increase primarily reflects investment in artificial intelligence. Alphabet kept its 2026 forecast at $195 billion to $205 billion and expects a significant increase in 2027. Microsoft expects about $170 billion in calendar-year 2026 spending, while Meta forecast $135 billion to $145 billion, including finance-lease payments.
Cash from operations did not cover quarterly equipment spending at three of the four companies, based on the figures provided. Amazon generated $42.4 billion, Alphabet $42.3 billion, Meta $32.8 billion and Microsoft $58 billion. The source author calculates that Microsoft had roughly $18.6 billion left after cash capital spending. The figures do not mean the companies cannot fund their plans; they show why external financing and contractual arrangements are becoming more important as AI data centers expand.
Recent financing moves illustrate the change. The source reports that Meta sold $25 billion in stock and $12.5 billion in preferred shares during the week, Microsoft issued $25 billion in bonds, and Amazon disclosed a loan facility of up to $80 billion secured by servers it plans to buy. The source also calculates that Amazon, Alphabet and Meta raised about $225 billion in long-term debt during the first nine months of 2026. These transactions take different forms and should not be treated as equivalent sources of cash or risk.
Infographic 01 · One quarter, four companies
The bill doubled. The cash did not.
What Amazon, Microsoft, Alphabet and Meta paid for buildings, servers and chips from July to September 2026, set against the cash their businesses brought in.
AmazonQ3 2025: $35.1B spent
Spent $34.9B more than came in
AlphabetQ3 2025: $24.0B spent
Free cash flow −$9.1B
MicrosoftQ3 2025: $19.4B spent
$18.6B left over
MetaQ3 2025: $18.8B spent
Free cash flow −$2.1B
Cash purchases of property and equipment as reported; not all of it is for AI (Amazon also builds warehouses). Gaps for Amazon and Microsoft are my arithmetic; free cash flow for Alphabet and Meta as the companies report it. Microsoft’s quarter is its fiscal Q1 2027 (July–September 2026).
Sources: Amazon results, 29 Oct · Microsoft 10-Q, 28 Oct · Alphabet results, 28 Oct · Meta results, 21 Oct. Debt raised: cash flow statements, nine months 2026.
Powered by Thorsten Meyer AI
Who Carries the Data Center Costs
The financing matters because building AI capacity involves more than buying servers. Companies also need land, buildings, networking equipment and electricity, often through arrangements that commit them to payments for years. The source’s review of four company filings puts their future payments on leases that had not yet started at about $1.02 trillion as of September 30, 2026. That is a calculated total of commitments, not an amount already borrowed or immediately due.
The time horizons can differ sharply. Servers are typically depreciated over five to six years, while data center leases may run as long as 20 years, power contracts 15 to 20 years, and some bonds for decades. If demand or AI revenue grows more slowly than expected, companies or their financing partners could still owe money on facilities and electricity. The source identifies newer GPU-cloud providers, building-owning side companies, lenders and investors as potential points where pressure could emerge. The size and distribution of any eventual losses remain uncertain.
For readers, the issue reaches beyond technology-company earnings. Investors and pension funds may hold debt or shares tied to the expansion, while long-term power demand can affect electricity systems and customers. Those are possible channels of exposure, not evidence that households will necessarily face a particular price increase.
Popular right now
Vitamix Propel Series 750 Professional-Grade Blender, 4 Programs, Black
De'Longhi Dinamica Plus Fully Automatic Espresso Machine
Dyson Ball Animal 3 Upright Vacuum
As an affiliate, we earn on qualifying purchases.
From Cash-Funded Growth to Mixed Financing
The source describes a shift from a period when large technology companies could fund data center construction largely from profits in advertising, software and online shopping. In 2026, the scale of planned spending has made financing a more visible part of the expansion. The figures cited are based on company filings and management guidance available by October 30, 2026; the combined totals are the source author’s calculations.
There are several ways to pay for the infrastructure. A company can use operating cash, issue bonds or take loans, lease facilities or equipment, work with partners that finance or own assets, or issue shares. Each route has different costs and obligations. Borrowing creates repayment and interest costs; leases create contractual payments; selling shares dilutes existing ownership; and partnerships can shift asset ownership while leaving the technology company with other commitments. Long-term power contracts add another layer of exposure.
Meta stated on October 26 that it intends to fund its AI investments “through a balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital.” That statement describes the company’s planned approach; it does not specify how much of future spending will come from each source.
Future Revenue Versus Long-Term Bills
The disclosures do not establish how quickly the companies will earn revenue from the new capacity, how fully the facilities will be used, or whether current spending forecasts will change. The roughly $1.02 trillion in future lease payments is a source-calculated figure for commitments on leases not yet started; the provided material does not break down the timing of every payment or the share tied specifically to AI.
It is also unclear how much risk sits with each technology company versus landlords, financing partners, lenders and investors. The source points to high interest rates on some debt issued by newer GPU-cloud providers, but the material provided does not give enough detail to assess their overall finances or compare their exposure directly with that of the four major companies. Claims about a potential “bubble” remain forecasts and judgments, not established outcomes.
Watch Spending and Financing Disclosures
The next indicators will be company results and filings showing actual capital spending, operating cash flow, new debt and changes in lease commitments. Investors will also be watching whether the companies maintain their 2026 forecasts and how they describe demand, utilization and returns from AI infrastructure. The source reports that Alphabet expects a significant spending increase in 2027, but no specific figure was provided.
Further information about partnerships, building ownership and long-term power contracts would help clarify how costs and risks are divided. Until then, the confirmed picture is a rapid rise in reported spending and a broader mix of financing—not a settled answer about whether future AI income will cover the commitments.
Key Questions
How much did the four companies spend in the third quarter?
Amazon, Microsoft, Alphabet and Meta spent a combined about $202 billion on property and equipment from July through September 2026, based on the source author’s calculation from company filings. The total was about $97 billion in the same period a year earlier.
Where is the money coming from?
The companies are using operating cash, bonds and loans, leases, partnerships and equity. The proportions vary by company, and the available source material does not provide a complete breakdown of funding for every project.
What does the $1.02 trillion lease figure mean?
It is the source author’s sum of future payments on leases that had not started as of September 30, 2026, across four company filings. It is a measure of contractual commitments, not a bill due immediately or debt already raised.
Does the spending show that AI data centers are a bubble?
No. Spending and financing figures alone do not establish whether the investment will earn adequate returns. The key unresolved issue is whether future revenue and demand will be sufficient to cover the long-term costs.
Source: ThorstenMeyerAI.com
