Rising Debt and AI Borrowing Could Put Government Bond Markets Under Strain
Government bond markets face growing pressure from rising inflation, high public debt and the rapid expansion of borrowing by artificial intelligence companies. A sharp, sudden jump in bond yields could spread losses across financial markets, especially if many investors try to sell at once.
Why Bond Yields Are Climbing
Yields on government bonds in several wealthy economies have reached levels not seen in many years. Higher energy costs linked to conflict in the Middle East have added to inflation worries. Investors also expect central banks may keep interest rates higher to slow price increases.
When bond yields rise, the market value of existing bonds usually falls. A sudden increase could also lower the value of other investments because investors use higher rates when judging what those assets are worth. This could affect artificial intelligence-related investments as well as government bonds used as loan collateral.
Public debt is close to its highest level since World War II and is expected to pass the size of the global economy. Heavy debt and the need to replace maturing loans make governments more exposed to sudden changes in borrowing costs, inflation and public spending.
AI Companies Add to Demand for Borrowing
Large technology companies and other firms building AI systems have raised hundreds of billions of dollars in debt. Their borrowing adds competition for investors’ money, which is also needed to buy government bonds.
Large cloud and technology firms now account for about 10% of global investment-grade bond sales, up from less than 1% in 2024. Much of the new AI-related borrowing is long-term. It is large enough to compete with government bonds for investment, even though it does not appear to be the main reason bond yields are rising.
Hedge Fund Trades Could Worsen a Selloff
Hedge funds now hold a larger share of U.S. government bonds than they did a few years ago. Their share rose from about 4% in 2022 to roughly 9% in 2025. Much of that increase came from highly leveraged trades, which use borrowed money to try to earn profits from small price differences.
Some of these trades have already been reduced without causing major problems in short-term lending markets. But borrowing remains high. If many funds hold similar positions and decide to exit at once, their sales could add to market stress.
The concern is that funds have both increased their use of borrowed money and built up trades that overlap. A shared rush to unwind those positions could make the effects larger and harder to contain.
Markets Are Not All in the Same Position
Trading has remained steady in government bond markets in the United States, Europe and the United Kingdom. Japan is a concern: trading there has become less liquid as bond yields have swung more sharply since the end of a policy that controlled the shape of interest rates.
Officials responsible for government borrowing have been urged to issue more long-term debt. Replacing long-term borrowing with large amounts of short-term debt can make it harder for a government to manage its loans when it must refinance them.
Connected AI Deals Could Carry Shared Risks
Financial ties between AI companies may create another source of risk. Fourteen publicly disclosed deals made since 2023 involved about $300 billion in equity investments, led by large cloud providers, and about $650 billion in debt for financially weaker AI firms.
In return, the investors received promises of about $1.4 trillion in future spending on services or computer chips. These commitments link the companies together: an investor may provide money to an AI firm and also expect that firm to spend money with it later.
If AI businesses do not earn enough from their products, several problems could emerge at the same time. The value of investors’ shares could fall, borrowers could struggle to pay their debts, and promised future purchases might not happen. That combination could create a triple hit for the companies involved and their investors.
The financing and spending promises are concentrated among a small number of AI firms. Trouble at one of these key companies could therefore spread through the connected deals and put pressure on other parts of the market.