Corporate borrowing to fund investments in artificial intelligence continues to accelerate, and bankers are responding by broadening the ways they place debt. The combination of massive spending on chips, cloud infrastructure and data centers has driven major technology firms - commonly referred to as hyperscalers - to turn to bond markets in currencies beyond the U.S. dollar to reach a wider investor base and reduce the risk of saturating the U.S. market with very large offerings.
Over the past 12 months, companies such as Amazon.com and Alphabet have issued about $60 billion of bonds across several currencies. "Alphabet and Amazon have diversified into other global markets in Europe, Canada, Asia," said Teddy Hodgson, global co-head of investment-grade debt at Morgan Stanley, describing the strategic shift toward non-dollar markets.
That move has set new benchmarks for corporate borrowing outside the dollar. Amazon's eight-part transaction in March raised €14.5 billion, which LSEG data shows was the largest ever corporate bond sale in the euro market. Alphabet's issuance likewise broke records across multiple currencies; its transactions in yen, Canadian dollars, Swiss francs and sterling each set borrowing records in those markets, according to LSEG. Alphabet also issued the first 100-year bond from a technology company since 1997.
Those large deals highlight the scale of funding requirements facing hyperscalers. BNP Paribas estimates capital expenditures for hyperscalers this year at roughly $725 billion - a figure that the bank says is nearly double the level seen in mid-2025. Analysts note that spending is rising faster than operating cash flow, which is contributing to a need for external financing.
New structures emerge for data center and AI financing
Alongside geographic diversification of bond issuance, banks are experimenting with innovative instruments to finance AI startups and data center operators. One approach that has gained traction is structuring debt around pre-arranged data center leases that can be committed before construction starts. These lease-backed deals are intended to provide investors with clearer visibility into future cash flows.
A recent instance of this approach was an $810 million note issued earlier this month by Stingray Compute, a unit of Cipher Digital. The offering was reportedly nine times oversubscribed, according to Cody Gunsch, head of North America leveraged finance capital markets at Morgan Stanley. The lending behind the transaction was secured by a data center lease to Amazon. Gunsch said the first deals with construction-loan-inspired structures began appearing last year, and about 15 of these offerings have since been sold to high-yield investors.
Stingray Compute did not respond to requests for comment regarding the transaction. Amazon said it regularly evaluates its operating plan and makes financing decisions - such as issuing bonds across currencies - accordingly. Alphabet pointed to remarks from CFO Anat Ashkenazi noting the company has accumulated $100 billion in outstanding debt across six major currencies, and to comments by CEO Sundar Pichai that investments are funded through a mix of cash flow, debt and equity.
Investor demand and concerns about market capacity
Despite the sharp increase in issuance to fund AI-related spending, demand from investors has remained robust. Bankers caution, however, that questions are emerging around the market's ability to absorb the sheer volume of new debt. Morgan Stanley's Hodgson said the scale of AI-related borrowing could push total investment-grade issuance above $2 trillion for the first time in 2026. Hyperscaler investment-grade deals have already exceeded their full-year 2025 total and are tracking toward BNP Paribas' forecast of $250 billion for the year.
"These are high-quality rated bonds, there is high appetite for them and there is a lot of liquidity around them in the market," said Victoria Fernandez, chief market strategist and fixed-income portfolio manager at Crossmark Global Investments, speaking about hyperscaler bonds. She added that repeated, frequent returns to the bond market by the same issuers could become a concern.
Some market participants are also taking note of recent stock offerings, reasoning that equity issuance could signal further capital needs and therefore a corresponding demand for debt. "If they are now issuing equity, how much more debt will they need to raise? These are questions that investors are working through," Hodgson said.
So far, there are limited signs of outright saturation. Barclays data indicates AI-related debt in the U.S. approaches 15% of total investment-grade issuance. Scott Schulte, global co-head of investment-grade debt syndicate at Barclays, observed that while the share of recent AI-related issuance relative to total issuance is sizable, it remains small when viewed against total outstanding debt within broader investment-grade credit indices.
Jeff Given, head of developed-market fixed income at Manulife Investment Management, noted that hyperscalers are still in the midst of ramping up long-term AI and data center investments, which helps keep the funding pipeline active. "As long as hyperscalers and data centers continue to be willing to spend more, demand is going to be there," he said.
Implications for markets and financing
The combined effect of multi-currency issuance and novel lease-backed debt structures is reshaping segments of the corporate bond market. Large, high-quality issuers are leveraging global demand to place historically large offerings in euros, sterling, yen and other currencies. At the same time, specialized credit products tied to data center operations are opening high-yield investor access to financing that aims to mirror construction lending protections and provide clearer cash-flow visibility.
Bankers and investors alike are watching whether demand will keep pace as issuance continues to grow. The next phase of this market evolution will depend on how hyperscalers balance capital spending, operating cash flows and the mix of debt and equity they turn to for funding.