Oracle caps new AI funding at $20B, targets $30B cash flow by 2029

Oracle is setting a limit on its borrowing or increments it will incur through investors for its AI growth. On October 28, 2026, the company declared that any new borrowing will be restricted to $20 billion until the end of financial year 2030. In addition, Oracle anticipates a minimum cash flow of $30 billion in fiscal year 2029.

At its Las Vegas investor day, Oracle sought to ease concerns about the cost of its massive AI buildout. The message was simple: its growing backlog of AI contracts should eventually bring in enough cash to help pay for the infrastructure needed to fulfill them.

The ceiling Oracle set, and the timeline behind it

Oracle’s finances won’t improve overnight. The company expects capital spending to reach its highest levels in fiscal 2027 and 2028. CFO Hilary Maxson said free cash flow should turn positive in the second half of fiscal 2028.

Oracle plans to raise no more than $20 billion in additional funding. That includes $10 billion from equity and another $10 billion in gross debt. The limit excludes refinancing and comes on top of the $20 billion share sale completed earlier in fiscal 2027.

Oracle anticipates that by the 2029 fiscal year, it will be able to sustain company growth without requiring any kind of new investment in the business.

The company also aims to ensure that the amount of its net debt does not exceed the equivalent of 3.5 times its adjusted EBITDA. These announcements pleased the investors, who on October 28 sent Oracle shares up by more than 12%.

Why the promise was necessary

Oracle has good reason to reassure investors. The company reported negative free cash flow of $23.7 billion in fiscal 2026. Another $5 billion went out in the first quarter of fiscal 2027.

Demand isn’t the problem. First-quarter revenue rose 30% to $19.3 billion, while cloud infrastructure sales surged 121% to $7.4 billion.

The trouble is that building data centers costs money long before customers start paying for the services. Oracle already had $125.34 billion in borrowings as of August 31, 2026.

Backlog is not revenue, and OpenAI is most of it

Oracle’s $664 billion backlog looks impressive, but that money isn’t in the bank. It represents future business that the company has yet to deliver. According to its SEC filing, roughly half should become revenue within 36 months.

A reported $300 billion computing deal with OpenAI, spread over five years, adds another concern. Oracle is counting heavily on one customer to fulfil a substantial part of its business.

To ease the financial burden, Oracle is turning to vendor financing, customer-supplied hardware and advance payments. Customers prepaid $11.36 billion of first-quarter capital spending, as Cryptopolitan reported.

What the AI math outside Oracle says

Oracle isn’t alone in facing tough questions about AI spending. J.P. Morgan expects global AI capital expenditure to reach $900 billion in 2026 and total $5.5 trillion by 2030. Companies are also relying more on debt and equity to finance their expansion.

But will AI generate enough revenue to justify those investments?

Bain estimates the industry could need nearly $6 trillion in annual revenue by 2031 to sustain its infrastructure spending. Existing consumer and enterprise applications would still leave a $4.2 trillion gap.

The stakes are high. On October 14, 2026, Gartner warned that more than $3.5 trillion in AI infrastructure spending could be at risk of write-downs by the end of 2030.

The numbers explain why. Gartner estimates that data-center operators would need $8.5 trillion in revenue through 2030 to recover their costs and earn a 12% return. But it expects them to generate no more than $3.6 trillion.

This is a challenge for Oracle’s promises. If it meets its objectives, that could confirm that large investments in AI can eventually be justified. If it misses its targets, it will only add to the burdens of a company swimming in debt and heavily dependent on OpenAI.