🔍 Read the full analysis: Are Investors Or Customers Paying For 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 reported about $202 billion in property and equipment spending in the third quarter of 2026, more than double the year-earlier total. Company filings and the source’s calculations point to a mix of operating cash, borrowing, leases, partnerships and share sales financing the build-out, with the costs and risks extending to investors and potentially electricity customers.
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment in the third quarter of 2026, more than twice their roughly $97 billion total a year earlier, according to a calculation from their cash-flow statements in the source report. The scale of the AI infrastructure build-out is pushing the companies to use borrowed money, leases and equity alongside operating cash, shifting some exposure beyond their own balance sheets.
Recent financing moves show how the spending is being funded. The source report says Meta sold $25 billion of its own stock and $12.5 billion in preferred shares during the week of October 26. Microsoft issued $25 billion in bonds, including debt with a maturity in 2066. Amazon disclosed a loan arrangement of up to $80 billion, secured by servers it will buy with the financing. These are distinct transactions and do not by themselves establish that each dollar will fund data centers alone.
The source’s quarterly spending totals are based on company filings: Amazon spent $77.3 billion, Alphabet $51.4 billion, Microsoft $39.4 billion in cash, and Meta $33.9 billion on property and equipment from July through September. Microsoft reported a higher figure of $45.2 billion when new finance leases are included. The companies’ stated full-year plans also point to continued investment: Amazon expects about $220 billion in cash capital spending in 2026; Alphabet forecasts $195 billion to $205 billion; Microsoft expects about $170 billion for the calendar year; and Meta projects $135 billion to $145 billion, including finance-lease payments.
Cash generation does not cover the same spending level at every company. The source calculates that Amazon, Alphabet and Meta each spent more on property and equipment during the quarter than their operating cash flow for that period. Microsoft generated $58 billion in operating cash against $39.4 billion in cash property and equipment spending. The report also totals about $1.02 trillion in future lease payments across the four companies as of September 30, based on commitments in their filings that had not yet started, adding to the capacity challenges facing data centers. Lease commitments are future obligations, not cash already spent.
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 Carries the Financing Risk
The financing mix matters because the infrastructure bill lasts beyond a single quarter. Investors who buy bonds or shares provide capital and take on risks tied to repayment, returns and company performance. Lenders may have claims against specific assets, as in Amazon’s server-backed loan. Leases and partnerships can place buildings or other assets on a different company’s balance sheet, but they do not make the underlying costs disappear.
The source report highlights a mismatch between asset life and payment commitments: servers may be 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 for 40 years. If AI services produce less revenue than anticipated, or demand changes before those obligations end, companies and their financing partners still have to meet the bills. The report identifies newer GPU cloud providers, building-owning side companies and their investors as possible pressure points; it does not establish that losses have occurred.
Electricity customers are another possible point of exposure, but the supplied material does not quantify whether or how data-center costs are being passed into power rates. That distinction is important: financing risk is not proof that customers are already paying more. Whether costs reach customers depends on local utility rules, contracts and infrastructure spending, none of which is detailed in the source.
Products Worth Considering
DJI Mic 3 Wireless Microphone System (2 Transmitters, 1 Receiver, Charging Case)
Anker SOLIX C1000 Gen 2 Portable Power Station
Bose QuietComfort Wireless Bluetooth Headphones with Active Noise Cancellation, Mic, USB‑C Charging, Deep Bass, 24‑Hour Battery – White Smok
As an affiliate, we earn on qualifying purchases.
How Big Tech Funds the Build-Out
For years, the largest technology companies could fund data centers largely from cash generated by advertising, software and online retail. The 2026 spending figures in the source indicate a much larger scale of investment. Companies are now combining five financing routes: operating cash, bonds and loans, leases, partnerships or separate asset-owning companies, and sales of equity. Long-term power contracts add another layer of commitments.
These routes distribute obligations differently. Operating cash avoids interest and repayment schedules but is limited by the money a business generates. Debt supplies funds immediately and requires repayment under its terms. A lease lets a company use a facility without buying it outright, while still creating future payment commitments. Equity raises capital by selling ownership interests, and partnerships can share the cost of buildings or equipment. The source report says Amazon, Alphabet and Meta raised about $225 billion in long-term debt in the first nine months of 2026, though that figure should not be read as debt devoted exclusively to AI.
Meta described its approach in an October 26 statement, saying it intended to finance AI investments through a “balanced mix” of operating cash flow, investment-grade debt, strategic partnerships and equity capital. That statement confirms the company’s intended funding mix; it does not provide a complete breakdown of the cost or risk of each project.
“through a balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital.”
— Meta, in a statement dated October 26, 2026
The Limits of Current Cost Data
The figures describe spending and financing commitments, but they do not show how much revenue the new infrastructure will generate or when. The source material also does not provide a project-by-project breakdown separating AI data centers from other property and equipment, even though companies have attributed much of the spending increase to AI.
The approximately $1.02 trillion lease total is the source author’s sum of four company filings; it represents future payments on leases not yet started, not a bill due immediately. The material does not detail the terms of every lease, power contract or partnership, nor how much risk counterparties have accepted. It also does not show whether electricity customers will bear higher costs. Those effects remain possible exposures, not confirmed outcomes.
Finally, the source frames the sustainability question as who holds the risk if AI revenue grows more slowly than obligations. That is a scenario, not evidence that the companies cannot pay. The major firms’ financial capacity, future demand, utilization of data centers and eventual returns will shape the outcome.
Company Spending Plans Ahead
The companies’ own forecasts point to more investment in the near term. Amazon expects about $220 billion in 2026 cash capital spending; Alphabet has maintained its $195 billion to $205 billion outlook and expects a significant increase in 2027; Microsoft expects about $170 billion for 2026; and Meta projects $135 billion to $145 billion for the year, including finance-lease payments. Those forecasts are company guidance, not guarantees of final spending.
The next useful evidence will come from subsequent quarterly filings and results: actual capital spending, operating cash flow, new borrowing, lease commitments and any changes to full-year plans. Investors and customers will also be watching whether AI-related revenue and data-center use grow quickly enough to support the commitments. Until those figures emerge, who ultimately bears the cost—the companies, their investors, financing partners or, in some places, electricity customers—remains partly unsettled.
Key Questions
Are customers directly paying for AI data centers?
The source material does not establish that customers are directly paying more for data centers. Companies are using cash, debt, leases, partnerships and equity, while any effect on electricity bills would depend on local utility arrangements and is not quantified here.
How much did the four companies spend in the third quarter?
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment from July through September 2026, according to the source author’s sum of their cash-flow statements. The category includes more than data centers alone.
What does the $1.02 trillion lease figure mean?
It is the source author’s total of future lease payments across the four companies for commitments that had not started as of September 30, 2026. It is a measure of future obligations, not cash already spent or an amount due all at once.
Why are companies borrowing and selling shares?
The reported scale of spending is outpacing quarterly operating cash flow at Amazon, Alphabet and Meta. Borrowing and equity sales add funding; leases and partnerships can also spread the cost or place assets with other parties.
Does the spending prove there is an AI bubble?
No. Spending and financing commitments show the size of the investment, but do not establish whether AI infrastructure is overvalued. The source identifies the central uncertainty as whether future AI revenue will support the long-term costs.
Source: ThorstenMeyerAI.com
