Morgan Stanley Predicts $570B AI Debt Issuance by 2026

• Morgan Stanley sees AI debt issuance topping $570B in 2026 amid data center and chip demand.
• Alphabet, Amazon, Microsoft, and Meta are estimated to spend about $700 billion on AI.
• Morgan Stanley predicts that annual spending by hyperscalers may top $1 trillion by 2027.

According to a new report from Morgan Stanley, global AI‑related debt issuance is set to top $570 billion in 2026, which would mark one of the biggest corporate borrowing sprees ever. The prediction reflects a notable demand for capital as tech firms scramble to build AI data centers and buy advanced chips.

Morgan Stanley says AI‑related debt had already hit about $236 billion by the end of May 2026, roughly four times the level in the same period last year. The company expects borrowing to speed up even more in the second half as hyperscalers continue pouring money into their investments.

The main drivers of this borrowing surge are the world’s largest tech companies. Morgan Stanley estimates that Alphabet, Amazon, Microsoft, and Meta will spend about $700 billion on AI‑related capital projects in 2026 alone. Going forward, annual spending by hyperscalers could top $1 trillion by 2027.

As such, to fund these huge projects, companies are tapping bond markets more and more, instead of relying solely on cash flow.

Importantly, they are not limiting themselves to the US dollar markets alone. Morgan Stanley notes that hyperscalers are also issuing debt in euros and other currencies, aiming to expand their investor base and tap into bigger pools of capital.

For bond investors, AI has become a major topic in global credit markets. On the other hand, for stock investors, this increase in borrowing shows just how determined Big Tech is to come out on top in the AI race.

Still, there are certain risks present. Last week, Morgan Stanley warned that AI demand is contributing to “chipflation.” Memory chip prices have shot up about sixfold in the past year as suppliers focus on high‑margin AI hardware. Those higher costs are now starting to affect everything from cloud services to consumer electronics.

Some analysts also worry that the sheer size of this borrowing could create concentration risks if AI spending ever slows down. Large banks are reportedly looking at ways to spread out parts of AI data center debt to a wider set of investors, which is a sign of how huge some of these financing deals are.

Related: Ray Dalio Warns AI Boom May Face Pressure Beyond Technology Risks