Tech giants including Nvidia, Oracle, and SpaceX are raising billions of dollars to advance AI infrastructure.

Wall Street has found a new way to bet on artificial intelligence (AI): by lending to companies building AI. Over the past several weeks, Nvidia (NASDAQ: NVDA) has issued $25 billion in bonds, Oracle (NYSE: ORCL) has planned to raise up to $50 billion this year, and Amazon (NASDAQ: AMZN) has borrowed over $80 billion since the beginning of 2025, as AI infrastructure spending has driven a broader financing wave. Nvidia's deal attracted over $85 billion in subscriptions, more than three times the company's offering size.

Now Elon Musk's SpaceX (NASDAQ: SPCX) wants to join in. According to reports, bankers are preparing a bond offering of at least $20 billion, just weeks after the rocket and satellite company's record initial public offering (IPO).

So, is all this borrowing a sign of confident, cash-rich companies financing their future—or the kind of excessive leverage that typically precedes trouble?

Nvidia's $25 billion offering this month is its first entry into the bond market since 2021, with demand far exceeding supply, leading the company to increase the deal from its initial target of about $20 billion. Amazon has been issuing debt denominated in various currencies, from euros to Canadian dollars, with its most recent offering being the largest corporate bond deal in that currency.

Oracle plans to split its financing between debt and equity, and has allocated approximately $40 billion for fiscal year 2027.

Meanwhile, Alphabet (NASDAQ: GOOG) (NASDAQ: GOOGL) chose a different route, raising approximately $85 billion through the sale of stock and preferred shares rather than debt.

But for companies choosing the debt financing route, is the level of debt concerning? In most cases, I would argue not.

For the largest borrowers, the data is reassuring. Nvidia alone earned approximately $43 billion in net income in its fiscal fourth quarter (ended January 25, 2026)—exceeding the entire $25 billion it just raised. Full-year revenue grew 65%, reaching approximately $216 billion. Companies that generate such steady profits don't need to borrow to survive. But it does create more options for the chipmaker.

Similar logic applies to Amazon and Alphabet. Both operate profitable cloud businesses that generate sufficient cash to cover debt interest. But more cash allows these companies to invest more aggressively in AI growth opportunities.

But there are some companies whose borrowing scale is concerning when compared to their underlying business.

For example, Oracle is profitable on paper, but its free cash flow in fiscal 2026 (ended May 31, 2026) was approximately negative $24 billion, as capital expenditures surged to nearly $56 billion. Its debt has now exceeded $100 billion.

SpaceX is the most glaring case, because it is simply not profitable. The company lost nearly $5 billion in 2025, and in the first quarter of 2026, with $4.69 billion in revenue, it lost $4.28 billion. Much of these losses come from its AI division, which it absorbed from Musk's xAI, with that division reporting $6.4 billion in operating losses on only $3.2 billion in revenue last year. Its rumored planned bond offering may not finance new growth, but rather refinance bridge loans coming due in 2027.

To be fair, SpaceX does have major contracts coming up. Google has agreed to pay $920 million per month from October 2026 through June 2029 for computing capacity, and there is a disclosed Anthropic arrangement that could total approximately $45 billion. Of course, both arrangements include termination rights.

The ongoing borrowing certainly deserves attention. But in most cases, it doesn't look like a major problem. At least not yet. Nvidia, Amazon, and Alphabet appear to be coming from a position of strength. However, Oracle's balance sheet looks more strained. That said, the company maintains meaningful profitable revenue despite its leverage.

On the other hand, SpaceX's potential bond offering is a bit concerning. However, I wouldn't view this trend itself as a warning sign. But cash-burning companies warrant closer scrutiny than cash-generating ones, because they truly need the capital, and they have the least margin for error if the returns on AI buildout take longer than the bond market expects.

Before buying Space Exploration Technologies stock, consider the following:

The Motley Fool Stock Advisor analyst team has just identified the 10 stocks they believe are most worthy for investors to buy right now…and Space Exploration Technologies is not among them. The 10 stocks picked could potentially generate massive returns in the coming year.

Consider Netflix, which was picked for this list on December 17, 2004…If you had invested $1,000 when we recommended it, you would have $417,305 now! Or Nvidia, which was picked on April 15, 2005…If you had invested $1,000 when we recommended it, you would have $1,293,148 now!

Now it bears noting that Stock Advisor's average total return is 936%—compared to 209% for the S&P 500, which is an overwhelmingly outperformance. Don't miss out on the latest top 10 list from Stock Advisor and join the investment community built by individual investors for individual investors.

*Stock Advisor returns as of June 20, 2026.

Daniel Sparks and his clients have no positions in any of the stocks mentioned. The Motley Fool holds positions in and recommends Alphabet, Amazon, Nvidia, and Oracle. The Motley Fool has a disclosure policy.