Technology companies’ growing demand for debt financing could help keep borrowing costs elevated across the wider economy.
The artificial-intelligence investment boom is becoming a significant force in bond markets, as technology companies borrow heavily to finance data centres and computing infrastructure. Analysts warn that this additional supply is competing with government debt for investors’ money, potentially putting upward pressure on long-term borrowing costs.
By August, US dollar investment-grade bond issuance had exceeded $1.5 trillion, according to reporting by Pound Sterling Live, putting 2026 on course to surpass the pandemic-era annual record. The publication cited an ING estimate that hyperscalers—the technology companies operating enormous cloud and computing networks—could issue $250 billion to $280 billion of debt this year. That represents a portion of the broader corporate borrowing total.
The financing requirements reflect the scale of the infrastructure buildout. In a September analysis, TwentyFour Asset Management estimated that AI-related capital expenditure would exceed $750 billion in 2026, with more than $1 trillion expected in 2027. These are forecasts, and the eventual spending totals will depend on how companies execute their investment plans.
Some companies are securing funding for exceptionally long periods. Amazon has issued dollar bonds maturing in 2076, while Alphabet’s sterling century bond matures in 2126. TwentyFour argues that these securities compete with long-dated government bonds for capital from pension funds, insurers and other institutional investors.
Fund managers have also highlighted the potential spillover. Speaking to Dow Jones, Hartford Funds’ Joe Boyle said the additional supply could keep yields elevated. Sit Fixed Income Advisors’ Bryce Doty explained that investors buying new technology-company debt may have to move money out of other parts of the bond market.
However, AI borrowing is one contributor to pressure on bond yields. TwentyFour also identifies government deficits, inflation and changing international demand for US Treasuries as major influences. Its analysis describes hyperscaler issuance as adding to those existing pressures.
The broader concern is that financing the AI expansion could make capital more expensive for other borrowers. Whether that happens will depend on the balance between new debt issuance and investor demand—and whether the enormous investments ultimately generate the returns their backers expect.
Sources:
