Artificial intelligence has become the defining investment theme of the decade, with hyperscalers committing hundreds of billions to data centers, chips, power, and software. The spending has rewarded companies across the AI supply chain, but markets are shifting from excitement to accountability.
Investors now want proof that massive capital expenditures can produce lasting returns. Banks and Goldman Sachs traders are warning that AI infrastructure spending is increasingly fueled by debt, while credit markets show growing concern as borrowing accelerates faster than near-term cash generation.
The AI arms race is rapidly reshaping balance sheets. Companies are issuing bonds at unprecedented scale to fund data centers, expand energy capacity, and secure scarce computing resources. The six largest AI infrastructure spenders — Microsoft (NASDAQ:MSFT | MSFT Price Prediction), Amazon (NASDAQ:AMZN), Alphabet (NASDAQ:GOOG), Meta Platforms (NASDAQ:META), Oracle (NYSE:ORCL), Nvidia (NASDAQ:NVDA), and SpaceX (NASDAQ:SPCX) — have issued a combined $244 billion in bonds this year, more than double last year’s total and 14 times 2024 levels.
This surge in borrowing reflects the enormous capital requirements behind AI. Building the infrastructure needed to support advanced models requires billions of dollars before those investments begin producing meaningful revenue. Unlike traditional software businesses, where additional users can often be added at minimal cost, AI requires expensive physical infrastructure.
Goldman Sachs reports that hyperscaler leverage ratios have doubled from 0.9x to 1.8x in roughly six months. Market indigestion is evident: Goldman’s AI bond basket spreads widened sharply, and the supply pain threshold has collapsed — where $75 billion once stressed the market, just $25 billion now puts it on the defensive.
The concern is not that AI lacks potential — it is that the timeline for turning massive investments into profitable businesses remains uncertain.
Credit default swap spreads have widened notably for major tech issuers, far outpacing the broader market. Investors are stepping back, wary of endless bond supply and questioning whether the market can absorb continued AI-related borrowing at the current pace. AI-related issuers now represent a growing and systemic share of investment-grade credit indices.