Artificial Intelligence Is Becoming the New Debt Bubble

AI sector quietly surpasses banks to become the most indebted sector in the market

Shares of Oracle jumped 25% after the company secured a promise of $60 billion annually from OpenAI—money that OpenAI has not yet earned, for infrastructure that Oracle has not yet built. Realizing it would require 4.5 gigawatts of capacity (equivalent to four nuclear reactors) as well as a new wave of massive debt. Currently, Oracle’s debt is 500% of its equity, compared with 50% for Amazon, 30% for Microsoft, and even less for Meta and Google.

This expansion cannot be financed from Oracle’s cash flows. The company will need to raise new debt or equity. Thus, the AI infrastructure boom, which so far has been funded by the profits of a few “hyperscalers,” is turning into a borrowed-money-driven race.

Massive investments financed by a future that doesn’t yet exist

The AI industry is “creating” hundreds of billions in investments that don’t actually exist—neither as profits nor as available capital. Most of the money comes in the form of debt.
Just to maintain the infrastructure, $500 billion in annual capital expenditures will be required, even though the main real revenue from AI today comes from $19.99/month subscriptions for students using chatbots.

According to Morgan Stanley, total spending on data centers will reach $2.9 trillion by 2028—$1.6 trillion for hardware and $1.3 trillion for buildings, energy, and maintenance. This means $900 billion in investment just in 2028, almost as much as the total capital expenditure of all S&P 500 companies in 2024.

This boom could temporarily add about 0.4 percentage points to U.S. GDP growth in 2025–2026. But the big question remains: who will foot the bill?

Credit markets take the hit

Morgan Stanley acknowledges that the capital requirements for AI are “staggering” and that credit markets—both public and private—will need to provide the bulk of the financing.
Of the expected $2.9 trillion, $1.4 trillion can be covered by cash flows, but there remains a $1.5 trillion shortfall that will need to be funded with new debt.

According to Morgan Stanley, the breakdown looks like this:

• $200 billion in corporate debt in the tech sector;
• $150 billion via securitizations (ABS and CMBS);
• $800 billion in private credit and asset-backed leasing;
• $350 billion from private equity, sovereign funds, and banks.

Thus, private credit will play a key role in financing the new AI infrastructure while interest rates remain high.

The credit bubble is already here

According to JPMorgan, AI-related companies now make up 14% of the Investment Grade Index, with total debt exceeding $1.2 trillion—more than the entire banking sector. This makes them the largest sector in the index, with corporate bonds of AI companies even trading at tighter spreads than the market average.

Among the largest debtors are Apple, Duke Energy, and Oracle, with Oracle raising the most concern due to its high leverage.

JPMorgan warns:

“If AI stocks experience a correction, the effect will spill over into credit markets. AI is now not just a market risk but a credit risk.”

When the bubble bursts

The main risk is that if AI expectations are not met, the market will face not just a stock correction but a debt crisis, because the infrastructure is financed with loans backed by future revenues that may never materialize.

In such a scenario, Oracle would be the first victim, and a debt collapse could spill into the real economy, as AI is already closely linked to the financing of energy and construction projects.

Is there a way out?

According to Bain & Co., the only hope is a technological breakthrough—such as quantum computers or more efficient AI chips that could reduce energy and computing costs. However, Bain warns that real breakthroughs are not expected for 10–15 years.

Until then, artificial intelligence remains not only a driver of innovation but also the main source of new global debt.

AI is no longer just a technological phenomenon—it is a financial bubble funded by record levels of corporate debt. The sector has surpassed banks in leverage and could become the next systemic threat to global markets if the expected profits fail to materialize.