Technology companies issued a record $108.7 billion in corporate bonds during Q4 2025, roughly double the amount from the prior quarter, according to Moody’s Analytics. For the full year of 2025, Big Tech firms collectively raised approximately $120 billion in corporate debt, almost entirely to fund artificial intelligence infrastructure.

Alphabet’s blockbuster deal sets the tone

The poster child for this trend landed in February 2026. Alphabet priced a $20 billion seven-part US dollar bond offering on February 9, upsized from an initial $15 billion target after investor demand went, to put it mildly, a bit wild.

The deal attracted over $100 billion in investor orders. That’s a 5x oversubscription ratio for a company that arguably doesn’t even need to borrow money. Alphabet achieved 25 basis points of spread compression during bookbuilding, meaning the cost of borrowing actually fell as more investors piled in.

Alphabet also introduced a 100-year sterling bond, the first time a tech company has issued century-duration debt since 1997.

The AI debt machine and its hidden risks

AI-related debt now accounts for approximately 15% of the entire US corporate bond universe. The so-called hyperscalers, think Alphabet, Microsoft, Amazon, and Meta, are the primary drivers. These companies carry pristine credit ratings and can borrow at razor-thin spreads above Treasuries. Credit spreads across the investment-grade market remain near historic lows, partly because so much of the new supply is coming from these well-capitalized issuers.

But the story gets more complicated once you move beyond the trillion-dollar names. Some newer AI and data center issuers are facing a very different reception from the bond market. Select credits in this space have recorded spreads as high as 3.75 percentage points above comparable bonds, a stark contrast to the tight pricing that Alphabet and its peers enjoy. Some of these bonds have already declined in price after issuance, suggesting that even the investors who bought in at launch are having second thoughts.

What this means for investors watching from crypto

When investment-grade credit spreads sit near historic lows, it signals that institutional capital is comfortable with risk. That risk-on posture tends to flow across asset classes, supporting everything from equities to digital assets.

Companies that borrowed $120 billion in a single year to build data centers are making a bet that AI revenue will arrive in sufficient scale to service that debt. If growth expectations don’t materialize, the resulting credit deterioration could tighten financial conditions broadly.

The 3.75 percentage point premium being charged to less established AI issuers, combined with post-issuance price declines in select bonds, suggests that the market is already quietly differentiating between AI winners and AI hopefuls.