The AI boom is starting to create a different kind of market question: not how much computing companies need, but how much debt investors are willing to provide to build it. Space Exploration Technologies Corp. (NASDAQ:SPCX) is becoming an unusually revealing test, with a reported $40 billion financing plan for Nvidia Corp (NASDAQ:NVDA) chips arriving only months after the company raised $25 billion through its first bond sale.

AI Is Getting More Expensive

SpaceX’s reported financing would include about $10 billion of bank loans and $30 billion of investment-grade debt. The proposed deal, expected to close in 2027, would fund purchases of Nvidia chips as SpaceX expands its AI infrastructure.

The scale matters because SpaceX has already demonstrated how quickly its financing needs can grow. The company raised $85.7 billion in net proceeds from its June IPO, according to its SEC filing, and then issued $25 billion of investment-grade senior notes later that month.

After that bond offering, SpaceX reported $38.4 billion of total debt as of June 30, compared with $22 billion at the end of 2025. Its June bond deal carried interest rates ranging from 5.35% to 6.65%, with a weighted average rate of 5.855%.

That does not mean SpaceX is in financial trouble. Quite the opposite: the company has access to investment-grade credit and enormous equity-market backing. But it does illustrate how quickly an AI infrastructure strategy can turn into a financing strategy.

The Debt Is Spreading

SpaceX is hardly alone.

Oracle Corp (NYSE:ORCL) said it expected to raise $45 billion to $50 billion during 2026 through a combination of debt and equity to expand Oracle Cloud Infrastructure, citing contracted demand from customers including Meta Platforms Inc (NASDAQ:META), Nvidia, OpenAI and others.

CoreWeave, Inc. (NASDAQ:CRWV), meanwhile, has built a financing model around borrowing against AI infrastructure and customer contracts. In August, the company closed a $2.6 billion loan facility backed by customer commitments.

The broader market is already noticing the shift. Neuberger Berman estimated that hyperscaler, data-center and semiconductor financing had reached roughly $165 billion before the midpoint of 2026, about $27 billion more than the entire 2025 total.

The implication is bigger than SpaceX. AI infrastructure is becoming one of the largest new sources of corporate borrowing, which means the next phase of the AI trade will depend not only on demand for chips and computing, but also on the willingness of credit investors to keep funding the buildout.

Investors Have a New AI Metric

For equity investors, that creates a less obvious metric to watch: the cost of capital behind AI growth.

A company can have enormous demand for computing and still face a different question if financing becomes more expensive. Higher borrowing costs can reduce the economics of data centers, chips and AI services before demand itself weakens.

SpaceX therefore offers an early test. If investors continue absorbing enormous AI-related debt at investment-grade rates, the industry’s spending cycle can keep expanding. If lenders begin demanding substantially higher compensation for that risk, the AI boom could encounter its first serious constraint not in computing capacity, but in the capital markets.

Watch what happens to the next wave of AI-related debt issuance — particularly its pricing, investor demand and credit ratings. SpaceX’s $40 billion proposal is significant, but the more important signal for stocks such as SPCX, NVDA, ORCL and CRWV may be whether Wall Street continues treating AI infrastructure as an investment opportunity worth financing at scale.

Photo: gguy / Shutterstock – ejk