AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

Before you orderOffer from Amazon

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
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment in July–September 2026, more than twice the amount a year earlier, according to company filings tallied by Thorsten Meyer AI. The companies are financing the build-out through operating cash, borrowing, leases, partnerships and stock sales, while the long-term costs and returns remain uncertain.

Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment from July through September 2026, more than twice their combined spending in the same quarter a year earlier, according to a tally of company filings by Thorsten Meyer AI. Their disclosures show how financing for the data center build-out is spreading across operating cash, borrowing, leases, partnerships and share sales, with costs extending well beyond the quarter in which equipment is purchased.

The $202 billion figure is the article author’s sum of the companies’ cash flow statements, not a separately reported industry total. It covers property and equipment, which includes data centers, servers and chips but is not limited to AI. Amazon also spends on warehouses. Amazon reported $77.3 billion in property and equipment purchases, Alphabet $51.4 billion, Microsoft $39.4 billion and Meta $33.9 billion. The corresponding total a year earlier was about $97 billion.

The article’s comparison with operating cash flow indicates that three of the four companies spent more on property and equipment than their businesses brought in during the quarter. Microsoft’s reported operating cash flow was $58 billion, leaving $18.6 billion after the $39.4 billion in purchases. Amazon’s reported purchases exceeded operating cash flow by $34.9 billion, based on the author’s arithmetic. Alphabet reported quarterly free cash flow of minus $9.1 billion and Meta minus $2.1 billion. These measures have different reporting bases, so they should not be treated as interchangeable.

The financing extends beyond the quarter’s cash spending. The source reports that Amazon, Alphabet and Meta raised about $225 billion in long-term debt during the first nine months of 2026. Microsoft sold $25 billion in bonds this week, with some bonds due in 2066. Meta sold $25 billion of its own stock and $12.5 billion of preferred shares, while Amazon disclosed a loan of up to $80 billion secured by servers it plans to buy. These are separate financing disclosures, and the source does not establish that every dollar raised is earmarked exclusively for AI.

At a glance
reportWhen: Third-quarter results and financing dis…
The developmentA review of third-quarter 2026 filings shows the four largest U.S. technology companies spent about $202 billion on property and equipment as they increasingly combine cash, debt, leases and other financing for data centers.
How To Follow The Money Behind AI Data Centers

Capital & commitments · Q3 2026

How To Follow The Money Behind AI Data Centers

Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment in July–September 2026. Their build-out draws on cash, debt, leases, partnerships and stock sales, while long-term costs and returns remain uncertain.

Amazon$77.3B

Property & equipment

Alphabet$51.4B

Property & equipment

Microsoft$39.4B

Property & equipment

Meta$33.9B

Property & equipment

01 / The scale

Where the quarter’s spending went

The total is a calculation from four cash flow statements, not a separately reported industry figure.

Q3 2026 property & equipment · USD

Amazon
$77.3B
Alphabet
$51.4B
Microsoft
$39.4B
Meta
$33.9B
Combined author tally$202B

Property and equipment includes data centers, servers and chips, but is not limited to AI. Amazon also invests in warehouses.

Cash flow comparison

Three companies spent more on property and equipment than their businesses brought in during the quarter, using different reported measures.

Microsoft

$58B operating cash flow left $18.6B after $39.4B in purchases.

Other reported measures

Amazon purchases exceeded operating cash flow by $34.9B by the author’s arithmetic. Alphabet reported free cash flow of −$9.1B; Meta, −$2.1B.

Operating cash flow and free cash flow have different reporting bases; they are not interchangeable.

02 / Funding routes

How the build-out is financed

Each route places costs and obligations with different companies, investors and partners.

01 · Internal cash

Operating cash

Profits from advertising, software and online shopping have funded data centers. Spending now exceeds operating cash flow for some firms.

02 · Borrowing

Bonds & loans

Amazon, Alphabet and Meta raised about $225B in long-term debt in the first nine months of 2026. Microsoft sold $25B in bonds; some mature in 2066.

03 · Secured credit

Server-backed loan

Amazon disclosed a loan of up to $80B secured by servers it plans to purchase.

04 · Future payments

Leases

The four companies’ future lease payments not yet started total about $1.02T as of September 30, per the author’s filing tally.

05 · Shared projects

Partnerships

Partners and separate property owners can distribute responsibilities for facilities and funding among several parties.

06 · Equity capital

Share sales

Meta sold $25B of common stock and $12.5B of preferred shares to support its investment approach.

These are separate disclosures; the source does not establish that every dollar raised is earmarked exclusively for AI.

“Through a balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital.”
Meta · statement dated October 26, 2026
“Primarily reflects investments in artificial intelligence.”
Amazon · October 29, 2026 results

03 / Time horizon

Assets may age faster than obligations

The source describes a mismatch in the duration of equipment and commitments. If demand or AI revenue grows more slowly than expected, some payments may continue after equipment loses value. That is a risk scenario, not proof losses will occur.

About $1.02 trillion in future lease payments

A contractual payment total across four filings—not an amount due immediately, current debt, or an AI-only commitment.

Servers
5–6 years
Power contracts
15–20 years
Data center leases
Up to 20 years
Some bonds
Up to 40 years

Bars compare approximate durations; some categories are ranges or maximum terms.

04 / What is unknown

Spending is not a return forecast

The filings show investment and commitments, but do not show how much revenue new data centers will generate, when it may arrive, or what utilization they will reach. The spending total alone cannot resolve the debate over a possible investment bubble.

AI revenue

No company-by-company estimate tied to this spending is provided.

Lease detail

The author’s $1.02T sum is not broken out by company or payment schedule here.

Debt use

The $225B covers three companies over nine months; it is not all financing or a direct use-of-proceeds statement.

Who bears risk?

Operators, lenders, bondholders, property owners and electricity customers may be exposed; amounts are not quantified.

05 / What to watch next

Follow spending and repayment plans

Coming filings and guidance updates can show whether purchases hold near the third-quarter pace, financing expands and lease commitments change.

1Filings

Track property and equipment purchases.

2Guidance

Compare plans with spending outlooks.

3Financing

Watch debt, equity and lease changes.

4Returns

Look for utilization and revenue evidence.

Company benchmarks

Amazon expects about $220B in cash capital spending for all of 2026, saying the increase “primarily reflects investments in artificial intelligence.” Alphabet kept 2026 guidance at $195B–$205B and expects a significant increase in 2027. These are company forecasts, not confirmed future outlays.

The cited reporting calendars differ: Microsoft’s July–September quarter is its fiscal first quarter of 2027; the other figures are presented as calendar Q3.

AI infrastructure · Capital, contracts & uncertainty Figures summarized from the source article’s tally of company filings and statements. Powered by Thorsten Meyer AI

Who Carries the Build-Out Costs

The financing matters because data centers require large commitments before their owners know how much revenue the new computing capacity will generate. The article’s tally puts future lease payments not yet started at about $1.02 trillion for the four companies as of September 30, 2026. That is the author’s sum of four quarterly filings; it is a contractual payment total, not an amount due immediately or a measure of current debt.

Different funding routes place obligations with different parties. A company paying from cash carries the immediate cost itself. Bonds create scheduled repayment obligations for the borrower and expose investors to its ability to pay. Leases commit companies to future payments, while partnerships and separate property-owning entities can distribute responsibilities among several parties. The article says the risk can reach investors, pension funds, companies that own data center buildings and electricity customers, although it does not quantify how much risk each group bears.

The timing of those commitments is a central concern. According to the source, servers are written off over roughly 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 demand or AI revenue grows more slowly than expected, companies may still owe money for facilities and power after the equipment has lost value. The source describes this as a mismatch in the duration of costs and assets, not proof that losses will occur.

Products Worth Considering

HP Sprocket Photo Booth Machine Instant Color Photo Printer with 10.1 Touchscreen, LED Ring Light, 250+ Frames, 3x4 ZINK Paper, QR Code Shar

HP Sprocket Photo Booth Machine Instant Color Photo Printer with 10.1 Touchscreen, LED Ring Light, 250+ Frames, 3×4 ZINK Paper, QR Code Shar

Logitech MX Master 4 for Mac, Ergonomic Bluetooth Mouse with Haptic Feedback, Ultra-Fast Scrolling, USB-C Charging, for MacOS, iPadOS, Windo

Logitech MX Master 4 for Mac, Ergonomic Bluetooth Mouse with Haptic Feedback, Ultra-Fast Scrolling, USB-C Charging, for MacOS, iPadOS, Windo

As an affiliate, we earn on qualifying purchases.

From Cash Funding to Mixed Finance

The source describes a change from a period when large technology companies could fund data centers from profits earned in advertising, software and online shopping. It says the scale of investment in 2026 has led them to use several financing methods at once. The filings cited cover different reporting calendars: Microsoft’s July–September quarter is its fiscal first quarter of 2027, while the other figures are presented as the calendar third quarter.

Company guidance gives some indication of the planned spending pace. Amazon said it expects about $220 billion in cash capital spending for all of 2026 and said the increase “primarily reflects investments in artificial intelligence.” Alphabet kept its 2026 capital spending guidance at $195 billion to $205 billion and expects a “significant” increase in 2027. Those are company statements about expected spending, not a confirmed final tally of future outlays.

Meta set out its approach in an October 26 statement, saying it intends to fund AI investments with a mix of operating cash flow, investment-grade debt, strategic partnerships and equity capital. The source also points to newer GPU cloud provider CoreWeave, whose senior notes it says carry interest rates of 8.5% to 9.75%. That example illustrates how financing terms can differ for companies beyond the four technology giants; it does not establish the borrowing costs of the sector as a whole.

““Through a balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital.””

— Meta, in a statement dated October 26, 2026, as quoted by Thorsten Meyer AI

Returns and Obligations Still Unclear

The spending and financing figures describe investment and commitments, but they do not show how much revenue the new data centers will generate or when. The source does not provide a company-by-company estimate of AI revenue tied to the spending, the expected utilization of new facilities or the returns investors may receive. The debate over a possible investment bubble therefore cannot be resolved by the spending total alone.

The approximately $1.02 trillion in future lease payments is an author-calculated sum from four filings. The source does not break out each company’s lease total here, show the payment schedule or establish which commitments relate specifically to AI rather than other operations. Similarly, the $225 billion debt figure covers borrowing by Amazon, Alphabet and Meta over nine months; it is not a complete measure of all financing or a direct statement of how the proceeds will be spent.

It also remains unclear how much of any shortfall would ultimately be borne by data center operators, lenders, bondholders, property owners or electricity customers. The source identifies those parties as potential holders of risk but does not quantify their exposure or report losses. Forecasts of future spending and revenue may change as companies update their plans.

Watch Spending and Repayment Plans

The next useful checkpoints are the companies’ coming quarterly filings and updates to their capital spending guidance. Those disclosures can show whether property and equipment purchases continue at the third-quarter pace, whether companies raise additional debt or equity, and how lease commitments change. Amazon’s stated 2026 cash spending outlook and Alphabet’s 2027 expectations provide benchmarks against which later company reports can be compared.

Investors and readers can also follow the details behind new borrowing: its amount, maturity, interest cost and any assets pledged as security. For leases and power contracts, payment schedules and contract terms will help clarify when obligations fall due. The source does not specify a single upcoming milestone that will settle whether the investment will pay off; the answer depends on both future demand and the costs companies have committed to carry.

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, according to Thorsten Meyer AI’s sum of their cash flow statements. The figure covers more than AI-related spending.

Does the $1.02 trillion lease figure mean that amount is due now?

No. The article describes it as the sum of future lease payments that had not started as of September 30, 2026. It does not give a payment schedule or say that the full amount is currently due.

What funding methods are the companies using?

The source identifies operating cash, debt, leases, partnerships and equity. Specific examples include Microsoft’s $25 billion bond sale, Meta’s stock and preferred share sales, and Amazon’s disclosed loan of up to $80 billion secured by servers it plans to buy.

Does the spending prove that AI data centers are a bubble?

No. Spending figures show the scale of investment, but the source does not provide enough information about future AI revenue, facility use or investment returns to establish whether the build-out will pay off.

Source: Thorsten Meyer AI

This content is for general information only and is not financial, tax or legal advice. Consult a qualified professional for decisions about your money.
You May Also Like

Nvidia, CoreWeave, and Nebius: Inside the Circular Financing of the GPU Boom

Exploring how Nvidia, CoreWeave, and Nebius are financing the GPU boom through circular investment models, with confirmed details and ongoing questions.

Steuer

Germany introduces updated tax rules affecting retirees, with potential for increased tax payments. Details remain under discussion, but the change impacts many seniors.

Aufstockung Von Drei Anleihen Des Bundes – Tenderergebnis

Bundesbank gibt Ergebnis der Aufstockung von drei Bundesanleihen bekannt, was auf eine verstärkte Kreditaufnahme des Staates hindeutet.

Christine Lagarde: Interview With Ouest-France

ECB President Christine Lagarde shares her views on the Eurozone economy in a recent interview with Ouest-France, highlighting challenges and future policy directions.