Major technology firms are pouring billions into AI infrastructure, financed increasingly through complex borrowing, raising questions about market stability as private credit and recycled investments grow amid rising default risks.
Microsoft, Meta, Amazon and Google are pouring vast sums into artificial intelligence, but the spending surge is increasingly being financed with borrowed money. That matters because the investment case depends on continued explosive demand for data centres, chips and cloud capacity. If the market cools, the pressure would quickly shift to earnings, valuations and debt servicing.
Morgan Stanley has estimated that as much as $2.9 trillion could flow into data centres and related hardware worldwide by 2028. The bank said global technology groups invested roughly $440 billion in 2025 and are already signalling about $740 billion for 2026. It also calculated that companies can fund around $1.4 trillion of that out of their own resources, leaving a financing gap of roughly $1.5 trillion.
That shortfall is increasingly being filled through debt markets. According to a report cited by Livemint, major technology groups including Google, Meta, Amazon, Microsoft and Oracle issued $121 billion of debt in 2025 alone, around four times their average annual borrowing over the previous five years. The money is being channelled into new data centres and AI infrastructure, with Meta’s planned site in Louisiana among the most prominent examples.
The financing structure is also more complex than a simple wave of corporate bonds. Morgan Stanley said about $800 billion of the expected funding could come from private credit rather than conventional banks or public markets. That makes the system less transparent for investors and supervisors, because these loans sit outside the traditional banking perimeter and are often distributed through funds backed by pensions and insurers.
There are also growing concerns about circular funding within the sector. Chipmakers and cloud providers are investing in firms that build or run AI systems, and those recipients may then spend some of the money on the same suppliers’ products. That can support demand in the short term, but it also means part of the market is being sustained by capital recycled within the industry rather than by end customers paying for AI services.
The risk is not necessarily an imminent crash, but a tightening of financial conditions if expectations prove too high. The article notes that default rates in private credit have risen in the US, even if the sector is still widely seen as better capitalised than traditional banks. Columbia University researchers have found private credit funds often operate with much thicker equity cushions than US banks, which may help absorb losses. Even so, a sharp reversal in AI spending could ripple well beyond technology shares and into broader credit markets.
Disclaimer: This content is intended for informational purposes only. Readers are advised to exercise their own judgement, conduct due diligence, or consult a qualified expert before acting on any information provided.
