🔍 Read the full analysis: Who Pays For AI’s Data Center Buildout? A Look At 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 the year-earlier total, according to figures compiled from company filings. The companies are funding the buildout through operating cash, borrowing, leases, partnerships and share sales; the scale of their future obligations raises questions about who would bear the costs if AI revenues lag.
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment from July through September 2026, more than double the roughly $97 billion they spent in the same quarter a year earlier, according to a sum of their company filings. The spending, much of it tied to data centers, servers and chips, is being financed not only from business cash but also through debt, leases, partnerships and share sales—a shift that could spread some of the AI buildout’s financial risk beyond the companies themselves.
The spending totals come from cash flow statements and filings, with accounting differences between companies. Amazon reported $77.3 billion in quarterly property and equipment spending, Alphabet $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 source’s figures show that each company spent more than it did in the year-earlier quarter.
The companies are combining several funding methods. Amazon, Alphabet and Meta raised about $225 billion in long-term debt during the first nine months of 2026, according to the source’s tally. Microsoft sold $25 billion in bonds during the week, including bonds with maturities extending to 2066. Meta sold $25 billion of its own stock and $12.5 billion in preferred shares. Amazon also disclosed a loan facility of up to $80 billion secured by servers it plans to purchase.
Future obligations are also substantial. The source’s compilation of four quarterly filings puts leases that have not yet started at about $1.02 trillion as of September 30, mostly for data centers. These are contracted future payments, not cash already spent or debt of the same kind as bonds. The reported figures suggest the buildout’s financing reaches across company balance sheets, landlords, lenders and investors.
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 AI Spending Reaches Beyond Big Tech
The financing mix matters because it affects who carries the cost if demand for AI services grows more slowly than expected. A company that pays from cash bears the expense directly. Borrowing creates payments to bondholders, while leases and partnerships can place ownership or financing obligations with other entities. The source argues that risk may also reach investors, pension funds and electricity customers, although the extent of any eventual losses or price effects is not established by the figures provided.
There is also a timing issue. The source says servers are typically written off over five to six years, while data-center leases can run up to 20 years, power contracts 15 to 20 years, and some bonds 40 years. If equipment becomes less useful or revenue falls short before those longer obligations end, companies or their financing partners may still owe money. Those differences in duration make the buildout’s return on investment a question for years, not just a single quarter.
The figures do not establish that the spending is unsustainable or that an AI investment bubble exists. They do show that the project is large enough to change how leading technology companies fund expansion—and that the answer to who pays is increasingly more complex than “the company’s profits.”
Products Worth Considering
EcoFlow DELTA 3 Classic Portable Power Station, 1024Wh, 1800W
Anker SOLIX S2000 Portable Power Station, 2,010Wh, 1,500W
Bose SoundLink Flex (2nd Gen) Portable Bluetooth Speaker
As an affiliate, we earn on qualifying purchases.
From Cash-Funded Campuses to Multiple Funding Sources
The source describes a change from a period when large technology companies could fund data centers from profits generated by advertising, software and online commerce. In 2026, the reported scale of investment has pushed firms to combine internal cash with loans, bonds, leases and outside capital. The source’s comparison of quarterly spending with operating cash flow finds that three of the four companies spent more on property and equipment than their businesses generated in operating cash during the quarter. Microsoft was the exception in that comparison.
The companies’ full-year plans underscore the size of the commitments. Amazon said it expected about $220 billion in cash capital spending for 2026 and said the increase primarily reflected AI investment. Alphabet kept its 2026 capital-spending guidance at $195 billion to $205 billion and forecast 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. These are company forecasts, not final spending totals.
Not every company involved has the same financing profile. The source points to newer GPU cloud providers such as CoreWeave, whose senior notes it reports carry interest rates of 8.5% to 9.75%, as one example of businesses facing different borrowing costs. It also identifies companies that own data-center buildings as part of the wider financing chain. Those examples do not establish that any particular borrower will default; they illustrate that the risks and costs differ across the industry.
“primarily reflects investments in artificial intelligence.”
— Amazon, in its reported third-quarter results
Returns and Long-Term Costs Remain Open
The spending and financing amounts are based on company filings and the source author’s calculations; they are not a single standardized measure across all four businesses. In particular, cash capital spending and spending that includes finance leases are not directly interchangeable. The $1.02 trillion lease figure covers future payments under leases that had not started by September 30, rather than an immediate bill.
It is not yet clear how much revenue the new data centers will generate, when they will reach full use, or whether AI demand will keep pace with the capacity being built. The source’s figures also do not establish what share of financing is ultimately borne by outside investors, property owners or customers, or how electricity costs may change. The companies’ spending forecasts may change, and the long-term obligations will be paid over different schedules.
The “bubble” question therefore remains unresolved. The figures document heavy investment and a wider range of financing sources; they do not, by themselves, prove that the investments will fail or that the companies cannot meet their obligations.
Watch Spending, Borrowing and AI Revenue
The next evidence will come from the companies’ subsequent quarterly filings and updates to their capital-spending plans. Readers can compare actual spending with the 2026 forecasts and watch whether operating cash flow continues to cover investment. Changes in bond issuance, lease commitments and financing arrangements may show whether the companies are relying more heavily on outside capital.
Revenue and utilization will be just as important as construction totals. Future disclosures may clarify how quickly new data centers are being brought online and how much income AI products and services generate. Until then, the confirmed picture is a buildout funded through a mix of company cash and obligations involving lenders, landlords, investors and other partners.
Key Questions
How much did the four companies spend in the third quarter of 2026?
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment from July through September, based on the source’s sum of company filings. The comparable total a year earlier was about $97 billion.
Who is paying for the AI data center buildout?
The reported financing comes from several sources: company operating cash, bonds and loans, leases, partnerships and stock or preferred-share sales. The mix varies by company, and the available figures do not show precisely how much risk each outside party ultimately carries.
What does the $1.02 trillion lease figure mean?
It is the source’s combined estimate of future payments under leases that had not started as of September 30, 2026, mostly for data centers. It is not cash already spent and should not be treated as equivalent to the companies’ current bond debt.
Do these figures prove that AI investment is a bubble?
No. They show rapid spending growth and substantial financing commitments, but do not establish whether future AI revenue will justify the investment or whether the companies will face financial distress.
What should readers watch next?
Upcoming filings can show whether companies meet or revise their spending forecasts, how operating cash flow compares with investment, and whether new borrowing and lease commitments rise. AI revenue and data-center utilization will also help indicate whether the investment is producing returns.
Source: ThorstenMeyerAI.com
