🔍 Read the full analysis: AI Data Centers On The Rise: Following The Money on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
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 figures compiled from company filings. The build-out is increasingly funded through debt, leases, partnerships and equity, although the long-term returns and who ultimately bears the costs remain uncertain.
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment from July through September 2026, more than twice their spending a year earlier, according to a calculation based on their filings. The surge, largely tied to data centers, servers and chips, is being funded not only from operating cash but also with debt, leases, partnerships and equity.
The quarterly total compares with about $97 billion in the same period of 2025, based on the source’s sum of the companies’ cash-flow statements. Amazon reported $77.3 billion in property and equipment spending, Alphabet $51.4 billion, Microsoft $39.4 billion in cash spending and Meta $33.9 billion. Microsoft’s figure rises to $45.2 billion when new finance leases are included, according to its filing.
New financing has accompanied the spending. The source reports that Meta sold $25 billion of its own stock and $12.5 billion of preferred shares, while Microsoft issued $25 billion in bonds, including debt maturing in 2066. Amazon disclosed it had arranged a loan of up to $80 billion, secured by servers it will buy with the proceeds. The source also calculates that Amazon, Alphabet and Meta raised about $225 billion in long-term debt during the first nine months of 2026.
Spending exceeded operating cash flow at three of the four companies in the quarter, according to the source’s arithmetic, underscoring the power bottleneck facing AI data centers. Amazon generated $42.4 billion in operating cash flow against $77.3 billion in property and equipment spending; Alphabet generated $42.3 billion against $51.4 billion; and Meta generated $32.8 billion against $33.9 billion. Microsoft reported $58 billion in operating cash flow and $39.4 billion in cash property and equipment spending. These comparisons cover the quarter and do not, on their own, establish the companies’ overall ability to fund future investment or manage data-center grid costs.
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 Is Funding the AI Build-Out
The financing mix matters because the cost of AI infrastructure does not stop at the initial purchase of servers. Debt requires repayment, leases create future obligations, and equity sales dilute existing ownership. Partnerships and companies that own data center buildings can also place some investment outside a technology company’s direct balance sheet, while leaving it with contractual payments or other exposure.
The source calculates that the four companies had about $1.02 trillion in future lease payments for leases that had not yet started as of September 30, 2026. That figure, compiled from four quarterly filings, is more than twice their combined long-term debt in the source’s comparison. It points to commitments beyond the spending recorded in a single quarter, although the total is not a measure of cash immediately due.
The central business question is whether AI-related revenue will grow enough to support the infrastructure and its financing costs. The companies have very different cash flows and balance sheets, so the spending figures do not imply equal exposure. If returns lag, the consequences could extend to lenders, property investors and other partners; the source also identifies electricity customers as a group that may face costs tied to power demand.
Products Worth Considering
DJI Mic 3 Wireless Microphone System (2 Transmitters, 1 Receiver, Charging Case)
EcoFlow DELTA 3 Classic Portable Power Station, 1024Wh, 1800W
Anker SOLIX C1000 Gen 2 Portable Power Station
As an affiliate, we earn on qualifying purchases.
From Cash-Funded Growth to Multiple Sources
The source describes a change from an earlier pattern in which large technology companies could fund data centers largely from profits generated by advertising, software and online commerce. In 2026, it says, the scale of AI investment has prompted companies to use several financing channels at once: operating cash, borrowing, leases, partnerships and share sales. That is a description of the financing mix, not evidence that any one method is new to the industry.
The companies’ own projections show how large the planned investment is. Amazon said it expected about $220 billion in cash capital spending for 2026, with the increase primarily reflecting AI investments. Alphabet maintained guidance of $195 billion to $205 billion and expected a significant increase in 2027. Microsoft expected about $170 billion for calendar 2026, while Meta forecast $135 billion to $145 billion, including finance-lease payments. The estimates use company-specific definitions and periods, so they should not be treated as perfectly comparable.
Long-lived infrastructure also raises a timing issue. The source says servers are typically written off over five to six years, while data center leases can run for as long as 20 years and some power contracts for 15 to 20 years. It also notes that some bonds mature over 40 years. The obligations can therefore outlast the useful life of particular equipment, though replacement cycles, contract terms and actual asset use vary.
Returns and Long-Term Exposure
The figures show spending and stated financing plans, but they do not establish how much revenue AI infrastructure will generate or when. The return on the new capacity remains uncertain, as does whether demand will grow quickly enough to justify the scale and timing of investment.
The source flags newer GPU cloud providers such as CoreWeave, whose senior notes it says carry interest rates of 8.5% to 9.75%, as potentially more exposed than the largest technology companies. It also points to building-owning entities and other investors. Those are risk assessments, not proof that any borrower will default or that losses will be passed on to pension funds or electricity customers. The full terms, counterparties and allocation of costs across individual projects are not provided in the material.
Company Spending Plans Ahead
The next evidence will come from subsequent company filings and capital-spending updates, which can show whether the current investment pace continues and how it is financed. Investors will also watch for changes to the companies’ 2026 spending guidance and Alphabet’s stated expectation of higher investment in 2027.
Future disclosures may clarify lease commitments, debt maturities, power arrangements and the revenue generated by AI services. Until then, the reported spending totals and financing plans show the scale of the build-out, but not whether it will earn returns sufficient to cover its long-term costs.
Key Questions
How much did the four companies spend in the third quarter of 2026?
Amazon, Microsoft, Alphabet and Meta spent about $202 billion combined on property and equipment from July through September, according to the source’s calculation from company filings. The same calculation puts their total at about $97 billion a year earlier.
Which companies are using financing beyond operating cash?
The source reports that Meta raised money through stock and preferred shares, Microsoft issued $25 billion in bonds, and Amazon arranged a loan of up to $80 billion secured by servers. The companies also use leases and strategic partnerships as part of their financing mix.
Does the spending show that AI data centers are unprofitable?
No. Spending more than operating cash flow in one quarter does not establish whether a company’s AI investments are profitable over time. The material does not provide a complete measure of AI revenue or project-by-project returns.
What is the reported $1.02 trillion lease figure?
It is the source’s sum of future payments for leases that had not yet started, mostly for data centers, disclosed in four companies’ filings as of September 30, 2026. It represents future contractual commitments, not an amount due immediately.
Source: ThorstenMeyerAI.com
