Oracle in the AI stack

The business and what the filings show

Oracle has turned itself into one of the largest AI infrastructure providers in a short time. Its cloud infrastructure revenue reached $22.1B over its latest four quarters, +94%, and it reports remaining performance obligations of $664B, the result of very large cloud contracts signed over the past year.

To serve that backlog Oracle spent $75.7B on capital projects against $46.9B of operating cash flow, leaving free cash flow of −$28.7B. It also discloses $260 billion of data-center leases signed but not yet started. Oracle expects to recognize only about 13% of its backlog as revenue over the next twelve months; most of it depends on capacity still being built.

Oracle is therefore the purest test in the stack of contracted demand versus funding risk. The backlog is real, but it is concentrated in a few very large contracts, some with customers that are themselves funded by investors. The cost and availability of the debt that bridges spending and revenue is the variable to watch.

How AI shows up in the numbers

Oracle’s AI exposure is concentrated in its cloud infrastructure business, which rents large GPU clusters to AI labs and enterprises. That business is still small relative to Oracle’s software revenue, but it is where almost all of the growth in backlog and capex comes from, so Oracle’s consolidated numbers increasingly reflect a handful of very large AI contracts.

Oracle’s $75.7B of capital spending over the latest four quarters is 13% of the $566B the five largest buyers spent over their latest four quarters. The five together raised capex +81% on the four quarters before; Oracle’s own revenue grew +22%.

How it talks about AI

In its latest earnings release, furnished to the SEC on September 10, 2026, Oracle mentioned AI 17 times in 2,775 words of narrative text, about 6.1 mentions per 1,000 words. Of its 11 sentences about AI, 3 included a number: a revenue figure, a growth rate, a customer count or another measurable result. Across the 8 releases we read, back to December 9, 2024, AI mentions per 1,000 words went from 5.2 to 6.1.

Specific claims are easier to check against the numbers than general enthusiasm, which is why the share of AI sentences that carry a figure is worth watching alongside how often AI is mentioned. The sentences below are quoted from the release as published, after removing safe-harbor text and company boilerplate.

• “Oracle booked more than $30 billion of additional AI cloud contracts in Q1 increasing its RPO to $664 billion.”
• “Since the end of Q4, Oracle also delivered more than 300,000 GPUs to its AI Cloud customers and almost triple the capacity delivered in Q4 FY26.”
• “Remaining Performance Obligations Customer demand for AI Cloud Training and Inferencing Services continues to grow faster than supply.”

Source: earnings release (Form 8-K, Exhibit 99.1), September 10, 2026.

The trend in the numbers

Since the start of 2023, Oracle’s largest quarter by revenue was Q3 2026 at $19.3B. The latest quarter set the revenue record for the period.

Quarter by quarter, Oracle’s revenue grew +14% in Q4 2025, +22% in Q1 2026, +21% in Q2 2026, +30% in Q3 2026 on the same quarter a year earlier. Growth has been accelerating.

Operating margin moved from 29% to 35% over the same span. Over the latest four quarters it turned 65% of its revenue into operating cash flow and spent 105% of revenue on capital projects, against 46% a year earlier.

Segments that matter for AI

The parts of Oracle’s reported results most tied to the AI build-out, from the segment and product disclosures in its 10-K and 10-Q filings.

Where it sits in the stack

Oracle sits in the AI clouds layer of the AI stack. It is paid by the software layer; its own spending flows to servers, networking, power & DCs, chips.

Among the 8 companies in this layer on our map, it ranks number 5 by revenue and number 5 of 7 by revenue growth. The full layer analysis is on the AI clouds page.

Latest four quarters for 10-Q filers; latest fiscal year for 20-F filers. Companies under $100 million of revenue are left out.

Against the 7 US-listed companies in its layer, Oracle grew +22% over the latest year versus +20% for the layer combined, and converted −40% of revenue into free cash flow versus 3% for the layer. It is growing roughly in line with its layer.

What has to go right, and what could go wrong

What has to go right. The very large cloud contracts behind Oracle’s backlog have to turn into revenue on schedule, and the customers behind them have to be able to pay, while Oracle funds the build-out at a cost of capital the contracts can bear.

What could go wrong. Oracle is the large buyer most dependent on financing: negative free cash flow, $260 billion of leases signed but not started, and a backlog concentrated in a few contracts. A customer that can’t pay, or more expensive debt, would hit it first.

These describe the conditions the business depends on, as the filings show them. They are not a view on the stock.

Through the lens of Investing in AI

Rob May’s book offers six mental models for telling AI moats from traps. The one most useful for Oracle’s layer is “Expect major market dislocations”: Most markets move slowly. AI does not. A single technical result can force the market to reprice an entire layer overnight. Assume fast repricings and sudden business-model breaks are normal, and position to benefit from shocks.

The question it asks of any company is: Which three to five assumptions have to hold for this investment to work? Applied to Oracle, the filings on this page are the place to start: the margin trend shows whether it holds scarce value or passes it through, the customer and financing disclosures show who it depends on, and the commitments show what it has promised before the revenue arrives.

Customers and concentration

US companies must disclose any customer that accounts for 10% or more of revenue, and foreign filers usually describe their largest customers. Concentration matters because a company’s growth is only as durable as the handful of buyers behind it, and in the AI stack those buyers are often the same few companies.

• “No single customer accounted for 10% or more of our total revenues in fiscal 2026, 2025 or 2024.” 10-K, June 22, 2026

Financing links

Investments, guarantees, purchase commitments, warrants and power contracts between Oracle and other companies, each quoted from the filing that discloses it. Numbers match the money-loop map.

• 21HPE → OracleCustomer warrant

  Warrant for up to 4.2M HPE shares, tied to networking in Oracle’s AI data centers

  “The Warrant was issued in connection with the Holder entering into a strategic collaboration with the Company to deploy the Company’s networking solutions in the Holder’s AI data centers.”

  HPE 10-Q for the period ended July 31, 2026, filed September 3, 2026

Commitments and backlog

• Backlog. Oracle reports $664B of remaining performance obligations, 9.3 times its revenue over the latest four quarters: contracted revenue not yet earned.
• Purchase commitments. $34.1B of unconditional purchase obligations as of August 31, 2026 (10-Q): what it has contracted to buy from its own suppliers.
• Server life. Oracle depreciates its servers and network equipment over 1 to 6 years, after changes it disclosed: 4→5 years (fiscal 2023); 5→6 years (fiscal 2025). See the useful-life calculator.
• Leases signed, not yet started. “As of May 31, 2026, we had $260 billion of additional lease commitments, substantially all related to data center arrangements, that are generally expected to commence between the first quarter of fiscal 2027 and fiscal 2029 and for terms of fifteen to nineteen years that were not reflected on our consolidated balance sheet as of May 31, 2026 or in the maturities table above.” 10-K
• Share of backlog due within a year. “Remaining performance obligations were $664 billion as of August 31, 2026, of which we expect to recognize approximately 13% as revenues over the next twelve months, 37% over the subsequent month 13 to month 36, 34% over the subsequent month 37 to month 60 and the remainder thereafter.” 10-Q

What to watch

Oracle’s next quarterly report will cover the quarter ending around November 30, 2026. Large companies must file a 10-Q within 40 days of a quarter’s end and a 10-K within 60 days of the fiscal year’s end, so the new figures should arrive by about January 29, 2027 at the latest. This page, the layer pages and the next quarterly report all update from that filing.

• Free cash flow and new borrowing each quarter, because Oracle is funding a build-out larger than its cash flow.
• Growth in cloud infrastructure revenue as backlog converts, the test of whether the contracts become revenue on schedule.
• The share of backlog due within twelve months, which shows how much of the demand is near-term rather than years away.
• Its place in the next quarterly AI Stack Report, which updates every company on the map from the new filings.

Figures from Oracle’s SEC filings, latest period ended August 31, 2026. Other companies: all company pages.