🔍 Read the full analysis: The Investors And Customers Funding AI Data Centers 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 by ThorstenMeyerAI.com. Companies are financing the build-out with operating cash, borrowing, leases, partnerships and equity, while the scale and duration of future obligations leave 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 the roughly $97 billion they spent in the same period a year earlier, according to a calculation from their cash flow statements by ThorstenMeyerAI.com. The surge is being financed not just from company cash, but increasingly through debt, leases, strategic partners and equity—arrangements that spread the costs and risks of the AI data-center build-out among companies, lenders, investors and, potentially, customers.
The report, published October 30, cites three financing developments this week: Meta sold $25 billion of its own stock and $12.5 billion of preferred shares; Microsoft borrowed $25 billion, including bonds with maturities extending to 2066; and Amazon disclosed a loan facility of up to $80 billion secured by servers it plans to buy. These are distinct transactions with different terms, and the source does not say that all the proceeds are exclusively for data centers.
Capital spending reported for July through September was $77.3 billion at Amazon, $51.4 billion at Alphabet, $39.4 billion at Microsoft and $33.9 billion at Meta. The figures are cash spending on property and equipment; Microsoft separately put quarterly capital spending at $45.2 billion when new finance leases are included. The companies also have different spending plans: Amazon expects about $220 billion in cash capital spending for 2026, while Alphabet’s guidance is $195 billion to $205 billion, Microsoft expects about $170 billion for the calendar year, and Meta expects $135 billion to $145 billion, including finance-lease payments, as demand for AI data-center compute capacity grows.
The report’s calculation compares spending with cash generated by operations in the quarter. Amazon generated $42.4 billion, Alphabet $42.3 billion, Meta $32.8 billion and Microsoft $58 billion. On that basis, three companies spent more on property and equipment than their operations generated; Microsoft was the exception. The source also estimates that the four companies had about $1.02 trillion in future lease payments for leases not yet started as of September 30, based on its sum of their quarterly filings, amid a rapidly expanding data-center build-out. That is a calculated total, not a single company disclosure.
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 Bears the Data-Center Bill
The financing mix matters because it determines who carries the cost if demand or AI revenue disappoints. Operating cash comes from the companies’ existing businesses. Bonds and loans put repayment obligations on their balance sheets; leases commit companies to recurring payments; equity sales share ownership with new investors; and partnerships can put some construction or ownership costs elsewhere. Each route has different risks and does not make the underlying resources free.
The time horizons also differ. The report says servers are generally written off over five to six years, while data-center leases may run for up to 20 years and power contracts for 15 to 20 years. Some bonds stretch to 40 years. If equipment becomes obsolete or AI income grows more slowly than expected, long-running commitments can outlast the assets or business case that prompted them. The report identifies potential exposure beyond the largest technology companies, including data-center property owners, specialist cloud providers, lenders, investors and electricity customers. How much risk each group ultimately bears depends on contract terms and future demand.
For customers, the immediate relevance is that AI infrastructure is being financed at a scale that may affect competition, power demand and the costs of providing computing services. The source does not establish that consumers will face higher prices or that any company will be unable to meet its obligations. It does show why monitoring financing and commitments—not just reported AI sales—can help explain the financial stakes of the build-out.
Products Worth Considering
Jackery Explorer 300 Portable Power Station, 292Wh
Jackery Explorer 1000 v2 Portable Power Station
Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Black
As an affiliate, we earn on qualifying purchases.
From Cash-Funded Growth to Mixed Finance
The source describes a change from the previous decade, when large technology companies could fund data centers from profits generated by advertising, software and online retail. In 2026, the pace of spending has made that approach less sufficient for several companies, according to its comparison of quarterly investment and operating cash generation. The report characterizes the shift as the use of five financing channels: operating cash, borrowing, leases, third-party ownership or partnerships, and new shares. Long-term power contracts add another kind of commitment.
Meta publicly described part of that approach on October 26, saying it intended to fund AI investment through a “balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital.” The report also points to a wider set of obligations than spending alone: the four companies’ estimated future lease payments, together about $1.02 trillion as of September 30. That number reflects payments due under leases that had not yet begun, according to the author’s calculation from the companies’ filings; it is not the same as current debt or cash already spent.
The financing is not limited to the four largest companies. The report names newer GPU cloud providers such as CoreWeave, whose senior notes it says carry interest rates of 8.5% to 9.75%, as an example of a company borrowing at higher rates. That comparison points to differences in financing costs, but the source does not provide a full market-wide accounting of data-center funding or a comparable risk assessment for every provider.
“Through a balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital.”
— Meta, as quoted in the October 26 statement described by ThorstenMeyerAI.com
Revenue Must Match Long Commitments
It is not yet clear whether AI revenue will grow quickly enough to support the planned spending and long-term obligations. The source lists investment estimates and financing arrangements, but does not establish future returns, the utilization of each new data center, or how much revenue will be attributable to AI infrastructure. Company guidance is forward-looking and may change.
The approximately $1.02 trillion lease figure is the author’s sum of four companies’ filings, and the source does not provide a full breakdown of when each payment will fall due or how the commitments are divided by facility and project. Nor does the available material quantify how much financing risk has been transferred to building owners, partners, lenders or investors. The report’s estimates of free cash flow and cash gaps should also be read in their stated periods: it cites Amazon’s trailing-12-month free cash flow at minus $42.4 billion, and quarterly figures of minus $9.1 billion for Alphabet and minus $2.1 billion for Meta.
The scale of spending alone does not prove that the investments will fail or that a financial bubble exists. Outcomes will depend on AI demand, operating income, energy availability, financing costs and the ability to put new capacity to productive use.
Watch Spending, Debt and Utilization
The next evidence will come from company filings and earnings updates showing whether capital spending continues at the current pace, how companies fund it, and whether AI-related revenue and cash generation rise alongside the investment. Investors and customers can track changes in debt issuance, lease commitments, capital-spending guidance and operating cash flow, while distinguishing cash purchases from spending counted through finance leases.
Further detail on the Amazon loan facility, its borrowing terms and the use of proceeds would clarify how much of the planned server investment is financed with secured borrowing. Updates on power contracts, new data-center capacity and its utilization will also help show whether the assets are generating enough activity to justify their long commitments. The source does not provide a date for a next financing milestone or establish what the companies will disclose next.
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 ThorstenMeyerAI.com’s sum of company cash flow statements. That compares with about $97 billion in the same period a year earlier.
Which companies were named in the spending comparison?
The comparison covers Amazon, Alphabet, Microsoft and Meta. The reported quarterly amounts were $77.3 billion, $51.4 billion, $39.4 billion and $33.9 billion, respectively. Microsoft’s figure rises to $45.2 billion when new finance leases are counted, according to the report.
What financing methods are companies using?
The report identifies operating cash, borrowing, leases, partnerships or third-party ownership, and equity sales. Long-term power contracts are another commitment associated with data-center development.
Does the spending prove that AI data centers are a bubble?
No. The figures show rapid investment and substantial financing commitments, but they do not prove that returns will fall short. Future AI demand, revenue, costs and asset utilization remain uncertain.
What is the $1.02 trillion lease estimate?
It is the report author’s combined estimate of future payments under leases that had not started as of September 30, 2026, based on four companies’ filings. It is not current debt or money already spent, and the source does not provide a full payment schedule.
Source: ThorstenMeyerAI.com
