🔍 Read the full analysis: AI Data Centers Have A Big Bill. Who Ends Up Paying It? on ThorstenMeyerAI.com
Get hardware and tech essentials 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 from July through September 2026, more than twice the amount a year earlier, according to figures compiled from company filings. The companies are funding the expansion with cash, borrowing, leases, partnerships and equity; who ultimately bears the risk depends partly on future AI revenue and the terms of long-term commitments.
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment in the third quarter of 2026, more than double the roughly $97 billion recorded a year earlier, according to a sum of their cash-flow statements by ThorstenMeyerAI.com. The expansion, driven largely by AI data centers, servers and chips, is being funded not only from company cash but also through debt, leases, partnerships and equity—arrangements that distribute some of the financial exposure among lenders, investors, landlords and potentially electricity customers.
Several financing moves this week illustrated the scale of the build-out. Meta said it sold $25 billion of its own stock and $12.5 billion of preferred shares. Microsoft borrowed $25 billion, with some of the bonds due in 40 years. Amazon disclosed a loan facility of up to $80 billion secured by servers it plans to buy. The source report says Amazon, Alphabet and Meta raised about $225 billion in long-term debt during the first nine months of 2026.
The quarterly spending totals were $77.3 billion for Amazon, $51.4 billion for Alphabet, $39.4 billion for Microsoft and $33.9 billion for Meta. Microsoft reported $45.2 billion when new finance leases are included. The figures cover property and equipment, not solely AI infrastructure; Amazon said its spending increase primarily reflected AI investments. Each company’s spending plans also point to continued outlays: Amazon expects about $220 billion in cash capital spending for 2026, while Alphabet forecasts $195 billion to $205 billion, Microsoft about $170 billion and Meta $135 billion to $145 billion.
The report’s author calculated that three of the four companies spent more on property and equipment in the quarter than their businesses generated in operating cash. Microsoft generated $58 billion from operations, compared with $39.4 billion in cash property and equipment spending. The author’s calculation puts Amazon’s trailing 12-month free cash flow at minus $42.4 billion, and quarterly free cash flow at minus $9.1 billion for Alphabet and minus $2.1 billion for Meta. These are reported calculations, not measures that establish whether the companies can meet future obligations.
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
How the Financing Spreads Risk
The financing mix matters because a data center creates obligations that can last longer than the useful life of the equipment inside it. The report says servers are typically written off over five to six years, while power contracts can run 15 to 20 years, some data center leases up to 20 years, and some bonds for 40 years. If the facilities produce less revenue than anticipated, the companies and their financing partners still have to manage payments under those agreements.
The report identifies a roughly $1.02 trillion total in future payments for leases that had been signed but had not started as of September 30, based on its sum of four company filings. Those commitments are not the same as current debt, but they indicate costs that may come due later. Risk can also sit with companies that own data center buildings, lenders, bondholders and investors in infrastructure ventures. Electricity customers may be affected by local power demand and the cost of expanding supply, though the source does not quantify any resulting bill increases.
Products Worth Considering

Apple MacBook Pro 2026 14.2 Laptop with M5 Pro chip, 15‑core CPU, 16‑core GPU, 24 GB Unified Memory, 1 TB SSD, Wi‑Fi 7 – Silver

Apple MacBook Pro 16.2-inch with M5 Pro, 18-core CPU, 20-core GPU, 24GB Memory, and 1TB SSD (Space Black)

Apple MacBook Pro 16-inch (M1 Pro, 10‑core CPU, 16‑core GPU, 16GB RAM, 512GB SSD) – Silver
As an affiliate, we earn on qualifying purchases.
A Record Build-Out Meets Financing
The report describes a change from an earlier model in which large technology companies generally paid for data centers from the profits of advertising, software and online shopping. Spending has risen to a scale where some companies are using several funding sources at once. Meta described its approach in a statement cited in the report as a “balanced mix” of operating cash flow, investment-grade debt, strategic partnerships and equity capital.
Those sources move obligations in different ways. Cash avoids interest payments but reduces funds available for other uses. Bonds and loans bring in money now in exchange for future repayment and interest. Leases let a company use a facility without owning it outright, while partnerships or separate building-owning companies can place some construction costs and financing outside the technology company’s direct balance sheet. Equity sales raise funds by selling ownership stakes, which can dilute existing shareholders. Long-term power contracts add another set of commitments.
The report also points to newer GPU cloud providers such as CoreWeave as potentially more exposed than the largest technology companies. It cites senior notes with interest rates of 8.5% to 9.75%. The rate alone does not show whether a borrower can repay; that depends on revenue, costs and the terms of its financing.
“through a balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital.”
— Meta, in a statement dated October 26 as quoted by ThorstenMeyerAI.com
Revenue Versus Long-Term Bills
It remains unclear how quickly AI services will generate revenue sufficient to support the infrastructure spending and long-duration commitments. The figures in the source report show outlays and financing plans, but do not establish the eventual returns on individual data centers, chips or power agreements. Lease commitments may also begin at different times, and the report’s $1.02 trillion total is an author calculation from company filings.
The ultimate distribution of losses, if projects earn less than expected, is also not settled. Contract terms, asset values, refinancing conditions and the performance of AI services will affect how costs are shared among technology companies, landlords, lenders, bondholders, shareholders and other parties. The source provides no estimate of how much, if any, of the infrastructure expense will reach household electricity bills.
Company Spending Plans for 2026
The next indicators will be company results and filings showing how actual spending compares with the full-year plans: Amazon’s roughly $220 billion cash capital-spending outlook, Alphabet’s $195 billion to $205 billion forecast, Microsoft’s estimate of about $170 billion, and Meta’s $135 billion to $145 billion range. Updated cash-flow figures, financing disclosures and lease commitments can show whether companies are continuing to rely more heavily on outside funding.
Investors and customers will also be watching for evidence that AI-related demand is translating into revenue, and for details of new data center, loan and power agreements. The available information does not establish whether the spending will deliver returns large enough to cover the long-term costs. Borrowing, equity and infrastructure investments carry financial risk, and the outcome remains uncertain.
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, according to a total calculated from their cash-flow statements. The comparable figure a year earlier was about $97 billion.
Are all of those costs specifically for AI?
No. The reported totals cover property and equipment, which can include more than AI facilities. The source says Amazon described its spending increase as primarily reflecting AI investments, but the totals are not presented as an audited breakdown of AI-only spending.
Who could bear the cost if AI data centers earn less than expected?
That depends on contracts and financing arrangements. Technology companies may face debt repayments and leases, while lenders, bondholders, shareholders and building owners may also bear financial exposure. Effects on electricity customers are possible in some circumstances, but the source does not quantify them.
What does the $1.02 trillion lease figure mean?
It is the source author’s calculation of future payments tied to leases that had been signed but had not started as of September 30, 2026, using four company filings. It is a measure of future commitments, not current debt due immediately.
Source: ThorstenMeyerAI.com
