Summary
The artificial intelligence boom is increasingly funded by borrowed money, but investors are starting to pull back. Major tech companies like Alphabet, Meta, Amazon, and Oracle have issued over $300 billion in bonds since the start of 2025. However, demand for these bonds is dropping fast, forcing companies to offer better terms to attract buyers. This shift could make it more expensive for AI companies to raise cash and may slow down the entire AI spending spree.
Main Impact
The key development is that investor demand for AI-related debt is falling just as companies ramp up their borrowing. For example, Amazon had to offer extra yield on a recent $25 billion bond sale because orders were only 2.5 times the bonds available, down from 3.2 times earlier this year. This shows that Wall Street is becoming less willing to provide the huge amounts of debt that AI companies need to build data centers and buy chips. If this trend continues, borrowing costs will rise, and AI companies may have to slow their spending.
Key Details
What Happened
Tech giants known as hyperscalers are borrowing heavily to fund AI infrastructure. Since the start of 2025, Alphabet, Meta, Amazon, and Oracle alone have issued more than $300 billion in bonds. Nvidia also sold $25 billion in bonds last month, its first such sale in five years. SpaceX, which now owns AI company xAI, sold $25 billion in bonds just days after its record IPO. The top five hyperscalers are expected to issue $300 billion annually in the coming years, up from $175 billion in 2026.
Important Numbers and Facts
Investor demand is dropping sharply. The cover ratio, which measures investor orders per dollar of bonds, fell from nearly 5 times in February 2026 to below 2 times in July. For comparison, the ratio for all investment-grade bonds only slipped by about half a point in the same period. Amazon had to offer 18 to 21 basis points of extra yield on its longest-term debt to attract buyers. JPMorgan estimates that hyperscalers will see $375 billion in debt proceeds from 2026 to 2030.
Background and Context
The AI boom requires massive amounts of money. Companies need to build data centers, buy expensive chips from Nvidia, and pay for electricity and cooling. They have been using cash from operations, selling stock, and borrowing from bond markets. But the bond market is becoming crowded. The U.S. government is also issuing a lot of debt to cover the federal deficit, which is on track to hit $2 trillion this year. This means AI companies must compete with the Treasury for investor money. If borrowing costs rise, it could make AI projects less profitable and slow down investment.
Public or Industry Reaction
Bank of America noted that investors are pushing back, saying the Amazon deal injected uncertainty into the AI supply outlook. Apollo Global's chief economist warned that falling cover ratios suggest investors may need wider spreads to absorb more hyperscaler debt. JPMorgan strategists said the widening is a rational response to the accelerating pace of issuance. The bearishness is also spilling over to the secondary market, where SpaceX's debt is now trading at levels similar to junk bonds. SpaceX stock has fallen 45% below its high and is trading below its IPO price.
What This Means Going Forward
If investor demand continues to drop, AI companies will have to offer higher yields to attract buyers. This will increase their borrowing costs and could force them to cut back on spending. The effects could ripple through the U.S. economy. AI-related investment has accounted for more than half of real GDP growth in recent quarters. Citi Research warned that if AI investment declines, it could cause a mild recession. Consumer spending, which has been supported by rising stock prices, could also fall if equity prices drop further. The rise of cheaper Chinese AI models, like Moonshot's Kimi K3, adds more pressure by threatening U.S. companies' revenue.
Final Take
The AI boom is facing a double problem. Not only are stock investors selling off, but bond investors are also losing appetite for AI debt. This could make it harder and more expensive for tech giants to fund their ambitious plans. If borrowing costs rise and spending slows, the entire AI industry may need to adjust its expectations. The days of easy money for AI may be coming to an end.
Frequently Asked Questions
Why are AI companies borrowing so much money?
AI companies need huge amounts of cash to build data centers, buy chips, and pay for energy. They borrow from bond markets because their own cash flow and stock sales are not enough to cover the billions they spend each year.
What does a falling cover ratio mean?
The cover ratio shows how many investors want to buy a bond compared to how much is available. A falling ratio means fewer investors are interested, so companies may have to offer higher interest rates to attract buyers. This makes borrowing more expensive.
How could this affect the economy?
AI investment has been a big driver of U.S. economic growth. If companies cut back on spending because borrowing costs rise, it could slow the economy. Citi Research warns it could even lead to a mild recession if AI investment drops sharply.
