🔍 Read the full analysis: How The AI Data Center Boom Is Financed 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 reported about $202 billion in combined property and equipment spending in the third quarter of 2026, more than twice the year-earlier amount. The companies are financing the build-out with a mix of operating cash, borrowing, leases, partnerships and new equity, while long-term commitments shift some risk to lenders, investors, landlords and power providers.
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, and are increasingly funding data centers with borrowing, leases, partnerships and equity as well as cash. The shift matters because the companies’ infrastructure commitments extend beyond the servers themselves, tying investors and other counterparties to costs that can last longer than the equipment is expected to be useful.
The combined spending figure is a sum of the four companies’ cash flow statements for July through September, according to the source analysis. Their reported outlays were $77.3 billion for Amazon, $51.4 billion for Alphabet, $39.4 billion for Microsoft and $33.9 billion for Meta. Microsoft’s total rises to $45.2 billion when new finance leases are included. The figures cover property and equipment, not AI spending alone, although Amazon said its increase primarily reflected investment in artificial intelligence.
Companies are drawing on several funding channels at once. Meta said it plans to finance AI investment through operating cash flow, investment-grade debt, strategic partnerships and equity. This week, Meta sold $25 billion of common stock and $12.5 billion of preferred shares, while Microsoft issued $25 billion in bonds, including debt extending to 2066. Amazon disclosed a loan arrangement of up to $80 billion secured by servers it plans to purchase. The source does not provide the full terms or drawdown schedule for that facility.
Cash flow figures help explain the financing mix. In the quarter, Amazon generated $42.4 billion from operations, Alphabet $42.3 billion, Meta $32.8 billion and Microsoft $58.0 billion, compared with their capital spending above. These comparisons do not by themselves show how each company funds all investment: businesses also have cash reserves, financing arrangements and other sources of capital. The source analysis says Amazon, Alphabet and Meta had negative free cash flow on the specified measures, while Microsoft remained positive.
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 Infrastructure Costs
Financing choices determine who bears the cost if AI revenues do not grow quickly enough to support the build-out. Debt puts repayment obligations on the borrowing company; leases commit it to future payments; partnerships and building-owning entities can place some investment outside a technology company’s direct balance sheet. Those arrangements do not eliminate the underlying cost. They distribute it among companies, lenders, landlords and investors.
The commitments are large relative to the reported debt, but they require careful interpretation. The source analysis estimates that the four companies had about $1.02 trillion in future payments on leases not yet started as of September 30, based on its sum of their quarterly filings. That is a reported estimate of future contractual payments, not a bill due immediately, and its scope is not the same as outstanding debt.
The timing mismatch is another exposure. The source says servers are commonly written off over five to six years, while data-center leases can run as long as 20 years, power contracts 15 to 20 years and some bonds 40 years. If demand or revenue falls short, a company or its counterparties could still face obligations after particular equipment has lost value or usefulness.
Products Worth Considering
Dell UltraSharp U2725QE 27-Inch 4K UHD IPS Black Monitor
FLEXISPOT EN1 Electric Standing Desk, 48 x 24 in, Black
Apple iPad Air (4th Generation, 10.9-inch, Wi-Fi, 256GB) – Space Gray
As an affiliate, we earn on qualifying purchases.
From Cash-Funded Builds to Mixed Finance
The source describes a change from an earlier model in which technology companies largely financed data centers from profits generated by advertising, software and online shopping. In 2026, reported infrastructure spending has risen sharply, and companies are layering multiple forms of financing rather than relying only on cash generated in the period.
Debt issuance is part of that change. The source estimates that Amazon, Alphabet and Meta raised about $225 billion in long-term debt during the first nine months of 2026. Microsoft’s $25 billion bond sale followed this week. The comparison illustrates increased use of borrowing, but the source does not give a company-by-company breakdown of the $225 billion estimate in the supplied material.
Leases and separate property-owning companies can make the funding structure less visible than a direct cash purchase or bond issue. Power supply is also part of the build-out: long contracts can secure electricity for data centers but may lock in costs well after servers have been replaced. The source frames the debate over whether investment is a bubble partly as a question of which parties hold those longer-term risks.
“through a balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital”
— Meta, in a statement on October 26, 2026
Returns and Contract Terms Remain Unclear
The scale of spending and commitments is documented in company results and filings, but the material does not establish how quickly the new infrastructure will generate revenue or whether returns will cover its full cost. Future demand for AI services, the pace of capacity use and the profitability of those services remain uncertain.
The source gives only limited details about some financing arrangements, including Amazon’s loan of up to $80 billion. It does not specify how much Amazon will draw, the complete borrowing terms, or how collateral and repayment work. Nor does it detail the ownership and guarantees behind all building partnerships and side companies. The $1.02 trillion lease estimate is the source author’s sum of four filings, rather than a single figure reported by the companies as a combined total.
The source also identifies newer GPU cloud providers such as CoreWeave as a potential pressure point, citing senior notes with rates of 8.5% to 9.75%. Those rates indicate borrowing costs for the cited notes; they do not, on their own, establish that the company cannot meet its obligations. The extent to which risks would reach investors, pension funds or electricity customers depends on contract terms and future business conditions not supplied here.
Watch Spending and Funding Disclosures
The next evidence will come from companies’ subsequent earnings reports, regulatory filings and financing announcements. Readers can compare planned capital expenditure with actual spending, operating cash generation, new borrowing and disclosed lease commitments to see whether reliance on outside financing continues to grow.
Company guidance in the source points to substantial investment continuing through 2026. Amazon expected about $220 billion in cash capital spending for the year; Alphabet maintained a range of $195 billion to $205 billion and anticipated a significant increase in 2027; Microsoft expected about $170 billion for calendar 2026; and Meta projected $135 billion to $145 billion, including finance-lease payments. These are company forecasts reported in the source, not final spending totals.
Subsequent disclosures may clarify how much of announced facilities is drawn, how long power and property agreements run, and which parties bear costs if projects are delayed or demand is weaker than expected. Until then, the reported spending and commitments show the scale of the financing challenge, but not the eventual return on the infrastructure.
Key Questions
How much did the four companies spend in the third quarter?
About $202 billion on property and equipment from July through September 2026, according to the source analysis’ sum of company cash flow statements. The amount was more than twice the roughly $97 billion reported for the same period a year earlier.
How are the companies paying for data centers?
The funding mix includes operating cash, bonds and loans, leases, partnerships and equity issuance. The companies do not all use each channel in the same way or to the same extent.
What does the $1.02 trillion lease estimate represent?
It is the source author’s combined estimate of future payments on leases, mostly for data centers, that had not yet started as of September 30, 2026. It is a future contractual commitment, not an amount due immediately or a combined debt balance.
Does rising borrowing prove the AI investment is a bubble?
No. The financing figures show how companies are paying for infrastructure, but they do not establish whether future AI revenue will justify the investment. Returns, demand and the allocation of contractual risk remain uncertain.
Source: ThorstenMeyerAI.com
