AI runs on debt: $570 billion in issuance and $3 trillion off the balance sheet
Morgan Stanley estimates global AI-related debt issuance in 2026 at nearly $570 billion. Big tech's off-balance-sheet obligations grew from $1.65 trillion to close to $3 trillion in two months.

AI financing no longer relies on profits. TMTPost, a Chinese tech outlet, cites Morgan Stanley estimates showing global AI-related debt issuance in 2026 approaching $570 billion, double the level of a year earlier.
Capital spending is even larger. The five largest technology companies plan to invest between $725 and $769.2 billion in 2026. Sequoia Capital estimates global AI infrastructure spending at roughly $1.5 trillion, while noting that the industry's profitability threshold has risen to $3 trillion.
Off-balance-sheet obligations are the most concerning metric. According to the cited data, they grew from $1.65 trillion in July 2026 to nearly $3 trillion in August. Cash flows reflect this change. In the second quarter of 2026, Alphabet recorded its first negative free cash flow, at minus $5.9 billion. Amazon also fell below zero.
A second Chinese-language source describes the same trend. Yicai published an interview with Bloomberg Economics' chief economist for Asia-Pacific. In 2025, capital spending by technology companies among the world's top 50 firms exceeded $500 billion, surpassing their cash reserves for the first time in history. In 2026, financing shifted to the bond market. Net borrowing by tech firms doubled compared to the previous year, far outpacing the scale of simultaneous IPOs.
The Bloomberg Economics economist describes this as a shift from a "savings glut" to an "investment glut" driven by AI and armaments. The macroeconomic impact is asymmetric. Rising supply of tech company bonds pushes up the long end of the yield curve, while monetary policy affects the short end. In the same month, the Fed raised rates by 25 basis points to a range of 3.75 to 4.00 percent, the first time in over three years.
Regulators are already seeing the risk. In September 2026, Japan's Financial Services Agency designated AI data center financing as a target for special audit for the first time. The focus is on the exposure of major banks and insurers to US projects. Typical projects use leverage of up to 90 percent.
Macroeconomic conditions are no longer favorable. The Federal Reserve raised rates by 25 basis points to a range of 3.75 to 4.00 percent, the first time in over three years. Two-year yields rose, and ten-year yields approached the 5 percent threshold. With infrastructure projects typically funded at up to 90 percent leverage, every basis point changes the profitability calculation.
The key question remains the obligations not visible on operators' balance sheets. They will determine whether the current cycle ends in a valuation correction or in lasting demand for computing capacity. Spending forecasts are impressive, but they do not set the limit of the boom. The profitability of a single project does.
Sources
2All figures and quotations in this text come from the sources listed below.
Content prepared by the editorial team with AI assistance.
Comments
0- No comments yet — be the first.