🔍 Read the full analysis: Who Pays When AI Data Centers Need More Power? on ThorstenMeyerAI.com
Get smart everyday buys 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 total a year earlier, according to figures compiled from company filings. The build-out is being funded through cash, debt, leases, partnerships and share sales, while the companies’ future lease and power obligations raise questions about who bears the cost if AI revenues lag.
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment from July through September 2026, more than twice the roughly $97 billion they spent a year earlier, according to a calculation based on their company filings. As the AI infrastructure build-out grows faster than some companies’ available cash, the firms are relying more on borrowing, leases, outside partners and new share sales—spreading the financial exposure beyond their own balance sheets.
The companies’ reported spending included data centers, servers and chips, though not every dollar was exclusively for AI. Amazon reported $77.3 billion in property and equipment spending for the quarter, 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 figures and comparisons with the previous year are drawn from company results and filings; the combined total is the source author’s calculation.
Recent financing announcements illustrate the range of funding. Meta said it sold $25 billion of common stock and $12.5 billion of preferred shares. Microsoft issued $25 billion in bonds, including debt due in 2066. Amazon disclosed a loan facility of up to $80 billion secured by servers it plans to purchase. The source reports that Amazon, Alphabet and Meta raised about $225 billion in long-term debt during the first nine months of 2026.
Companies are also committing to pay for facilities over time. The source’s sum of four quarterly filings puts future payments on leases that had not yet started at about $1.02 trillion as of September 30, mostly for data centers. That is a reported estimate of contractual future payments, not cash already spent or debt currently due.
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 Infrastructure Costs Spread
The funding mix matters because it determines who carries the risk if demand for AI services or revenue growth falls short of expectations. A company that uses cash bears the cost directly. Bondholders and lenders take credit risk; landlords and infrastructure partners invest in facilities and depend on tenants’ payments; shareholders face dilution when a company sells equity. Electricity customers may also be affected if new data-center demand leads to costly grid upgrades or power arrangements, though the source does not quantify those effects.
The commitments also last longer than some of the equipment they support. The source says servers are generally written off over five to six years, while power contracts can run 15 to 20 years, data-center leases up to 20 years, and some bonds 40 years. If hardware becomes outdated or AI income grows more slowly than expected, companies and their financing partners may still owe payments long after the equipment’s useful life. That mismatch is a risk to monitor, not proof that losses will occur.
The major technology companies have substantial businesses beyond AI, but exposure differs across the supply chain. The source points to newer GPU cloud providers, including CoreWeave, whose senior notes carry reported rates of 8.5% to 9.75%, as a potentially more exposed part of the market. The available material does not establish that any borrower is unable to meet its obligations.
Products Worth Considering
Apple MacBook Pro (14.2-inch, 2025) – M5 chip, 16GB Unified Memory, 1TB SSD, Space Black
Samsung Essential S3 27-Inch Curved FHD Monitor
FLEXISPOT 42-Inch Electric Standing Desk Converter
As an affiliate, we earn on qualifying purchases.
From Cash-Funded Growth to Borrowing
For years, the largest technology companies could fund much of their infrastructure from cash generated by advertising, software and online commerce. The 2026 spending figures indicate a larger funding need. Amazon said it expected about $220 billion in cash capital spending for the year, primarily reflecting AI investments. Alphabet maintained a $195 billion to $205 billion capital-spending outlook and forecast a significant increase in 2027. Microsoft expected about $170 billion in calendar-year 2026 spending, while Meta projected $135 billion to $145 billion, including finance-lease payments.
Cash flow figures help explain why financing methods are expanding. The source reports that three of the four companies spent more on property and equipment during the quarter than their businesses generated in operating cash flow. Its calculations put Amazon’s free cash flow over the previous 12 months at negative $42.4 billion, Alphabet’s quarterly free cash flow at negative $9.1 billion, and Meta’s at negative $2.1 billion. These measures cover different time periods and should not be treated as directly equivalent. Microsoft remained clearly positive in the source’s quarterly comparison.
Meta described its plan on October 26 as a mix of operating cash flow, investment-grade debt, strategic partnerships and equity capital. That statement captures the broader change: companies are not choosing one funding source, but combining several. The figures in this report come from the source author’s aggregation of company filings and announcements; they are not independently audited as a combined industry total.
““Through a balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital.””
— Meta, in a statement dated October 26, 2026
Uncertain Returns and Long Commitments
The reported spending totals do not show exactly how much went to AI infrastructure rather than other equipment and facilities. Nor do the filings summarized in the source establish what share of future lease commitments is tied specifically to AI, how much of the planned power is secured, or what electricity and grid upgrades will cost. The roughly $1.02 trillion in future lease payments is a combined estimate from the source author, and its timing depends on leases that had not begun by September 30.
It is also unclear whether revenue from AI services will grow quickly enough to cover the full cost of equipment, buildings, financing and power over time. The source describes a risk of a mismatch between asset life and contractual obligations; it does not show that the companies will fail to pay or that an AI investment bubble has been established. Actual exposure will depend on contract terms, future demand, resale or reuse of equipment, and how much infrastructure is shared with other customers.
Watch Spending and Financing Updates
The next evidence will come from companies’ subsequent earnings reports, filings and financing disclosures. Readers can compare actual capital spending with each company’s 2026 guidance, track changes to its debt and lease commitments, and watch whether operating cash flow keeps pace. Updates to power agreements and data-center construction plans may clarify how much infrastructure is committed and who is responsible for its cost.
For now, the confirmed picture is a major increase in spending and a broader set of funding sources. The distribution of long-term risk—and whether AI income will justify the build-out—remains a question for future results rather than something settled by the spending figures alone.
Key Questions
Who is paying for the AI data-center build-out?
The companies themselves, lenders, shareholders and infrastructure partners are all funding parts of it. The four firms are using operating cash, bonds and loans, leases, partnerships and share sales.
How much did Amazon, Microsoft, Alphabet and Meta spend in Q3 2026?
They spent about $202 billion on property and equipment from July through September, according to the source’s sum of company filings. The total includes spending beyond AI alone.
What does the $1.02 trillion lease figure represent?
It is the source author’s estimate of future payments on leases that had not started as of September 30, 2026, based on four company filings. It is not cash already paid or an amount all due immediately.
Could electricity customers bear some of the cost?
They could be affected if data-center growth leads to higher power costs or grid investment that is passed through to customers. The source material does not quantify those effects or establish how costs will be allocated.
Does the spending prove there is an AI bubble?
No. The figures show rapid investment and long-term financial commitments, but they do not establish whether future AI revenue will justify the spending. The scale of returns and how risk is distributed remain uncertain.
Source: ThorstenMeyerAI.com
