As Big Tech borrows record amounts to fund an unprecedented infrastructure buildout, economists warn the borrowing binge could pose a systemic threat to the global financial system.
NEW YORK - The race to dominate artificial intelligence has entered a precarious new phase, shifting from software breakthroughs to massive financial liability. According to new data released in December 2025, major technology companies have issued over $120 billion in debt this year alone to fund AI infrastructure, a staggering sum that economists warn could threaten broader financial stability. The surge marks a departure for a sector traditionally known for its cash-rich balance sheets, raising alarms about a potential credit crunch reminiscent of past market bubbles.
Reports from Reuters and Fortune indicate that the world's largest AI companies-including Meta, Amazon, Nvidia, and Alphabet-have aggressively tapped into bond markets to finance the construction of data centers and the acquisition of high-performance chips. This borrowing spree has pushed global bond sales to a record $6 trillion in 2025, with the tech sector acting as a primary driver. As interest rates remain a critical factor, the scale of this leverage is drawing sharp scrutiny from market regulators and veteran economists.
Table of contents [Show]
The velocity of debt issuance accelerated dramatically in the final quarter of 2025. Data from Bank of America cited by Reuters reveals that $75 billion of investment-grade debt was issued by AI-focused Big Tech firms in September and October alone. To put this in perspective, the sector's average annual issuance between 2015 and 2024 was merely $32 billion. In just two months, these companies more than doubled their historical yearly average.
Goldman Sachs analysis paints an even starker picture, suggesting that companies tracked in their AI equity basket have accounted for $141 billion in corporate credit issuance year-to-date, surpassing the total debt raised in all of 2024. This capital is being funneled directly into physical infrastructure. Specific U.S. secured debt tied to data centers skyrocketed 112% to $25.4 billion in 2025, as operators race to build facilities capable of handling the immense power requirements of next-generation AI models.
The nature of this debt is becoming increasingly complex. In October, Meta concluded a $27 billion financing deal with Blue Owl Capital to fund its largest data center project to date. Meanwhile, OpenAI's ecosystem is heavily leveraged; reports indicate that partners tied to OpenAI hold approximately $100 billion in bonds, loans, and private credit-a figure equivalent to the net debt of global giants like Volkswagen.
The sheer magnitude of borrowing has elicited stern warnings from top financial experts. Mark Zandi, chief economist at Moody's Analytics, recently described the trend as a "mounting potential threat to the financial system." In a LinkedIn analysis referenced by Fortune, Zandi noted that the current leverage eclipses the borrowing seen prior to the dot-com crash, raising fears of a similar correction if AI revenue fails to materialize quickly enough to service these obligations.
"Hyperscalers are taking on high levels of debt through record bond issuance in 2025, eclipsing the borrowing by tech companies in the years before the dot-com bubble popped." - Mark Zandi, Chief Economist, Moody's Analytics
Market indicators are already reflecting this anxiety. Oracle, a key player in the cloud infrastructure race, has seen the cost of its five-year credit default swaps-insurance against default-shoot to record highs, according to Reuters. This signals that investors are beginning to price in significant risk, despite the company's strong market position.
The implications of this debt binge extend beyond corporate balance sheets. Forbes analysts suggest that a "trillion-dollar AI borrowing binge" could spark a credit crunch, potentially crowding out other sectors from accessing capital. With telecom giants like Verizon and AT&T also holding massive debt loads, the corporate bond market is becoming increasingly saturated.
Technologically, this aggressive spending confirms that the AI "arms race" is capital-intensive rather than code-intensive. Success is now predicated on the ability to finance hardware at a scale that excludes smaller competitors, potentially cementing a monopoly for the few hyperscalers capable of raising $100 billion at will.
Despite the warnings, the spending shows no signs of slowing. Citi analysts estimate that robust AI demand will drive hyperscalers to spend $490 billion on infrastructure in 2026 alone-$70 billion more than previous forecasts. Furthermore, projections from Sage Advisory indicate that AI capital expenditure could reach $600 billion by 2027.
As 2025 closes, the tech industry has effectively bet the house on AI. The next 12 to 24 months will be critical: companies must prove that this expensive infrastructure can generate proportionate profits. If they succeed, the debt will be viewed as a smart lever for growth. If they fail, the financial system may face a reckoning rooted in silicon and speculation.
An Indian tech CEO shares insights from 25 years of experience, arguing that AI is an amplifier, not a replacement. The article explores 10 real-world case studies where human empathy, creativity, and strategic thinking remain irreplaceable in business and society.
An Indian tech CEO argues that AI is an amplifier, not a replacement for humans. The article explores 10 real-world case studies, from medicine to leadership, where human empathy, creativity, and strategic thinking remain fundamentally superior to algorithms.
An in-depth look at why AI is a tool for human augmentation, not replacement. This article explores 10 case studies where human skills like empathy, strategic thinking, and ethical judgment remain superior, arguing for a future built on human-AI collaboration.
