The $570 Billion AI Debt Question: A Bubble Bigger Than the Dot-Com Era?

The pace of AI-related borrowing has entered unprecedented territory. In June 2026, Man Group, the world’s largest publicly traded hedge fund firm, warned that "bubble risks" are mounting in credit markets as bond sales used to fund the artificial intelligence buildout shatter records . The firm's argument is not that AI lacks transformative potential, but that the financing cycle is expanding far faster than any credible adoption curve can justify .

The numbers driving this warning are staggering. Morgan Stanley forecasts global AI-related debt issuance will more than double to nearly $570 billion in 2026, with $236 billion already sold by the end of May—a fourfold increase from the same period in 2025 . For context, internet and tech bond issuance during the dot-com era peaked at $85 billion in 2001, meaning the current AI wave is on track to dwarf that figure by a factor of nearly 7x .

Nvidia’s Record Bond Sale Shows the Frenzy

The demand for AI credit was laid bare on June 15, 2026, when Nvidia sold $25 billion in high-grade bonds, its first corporate debt offering since 2021 . The deal drew roughly $85 billion in orders, more than three times the size of the bond, allowing the chipmaker to upsize the offering from an initial $20 billion target .

The sale, structured across seven tranches with maturities up to 2056, was less about an immediate cash crunch and more about establishing a liquid benchmark in investment-grade debt markets to fund the capital-heavy expansion behind its AI hardware dominance . Nvidia’s move mirrors a broader trend. Major tech firms are turning to debt markets on a massive scale to fund AI infrastructure and data centers without issuing additional equity and diluting shareholders .

Where the Credit Risks Are Concentrated

Man Group’s warning points to specific pockets of vulnerability within the broader AI debt boom.

Cash-Flow-Negative Borrowers in High-Yield Markets

The firm is most concerned about the high-yield and leveraged loan markets, where many AI borrowers are free-cash-flow negative . These companies are borrowing to sustain operations and capital spending without generating enough cash to service their existing debt, a classic precursor to default waves when market conditions tighten .

Doubling Leverage at Hyperscale Cloud Companies

Leverage ratios at hyperscale cloud companies have surged from 0.9x to 1.8x in just two quarters, with capital expenditure growth continuing to outpace revenue and free cash flow growth . The full impact of depreciation pressures from the massive infrastructure buildout has yet to materialize, meaning leverage metrics may continue to deteriorate .

Passive Bond Fund Exposure

AI-linked debt has already surpassed US banks as the largest investment-grade market segment, a shift that occurred by October 2025 . This means passive bond funds, which track indices and cannot choose to avoid AI exposure, are now heavily weighted toward a sector where credit risk is building .

Private Credit Stress

Beyond the public markets, private credit firms are seeing a series of high-profile defaults on AI and software deals . Opaque financing structures, such as the use of special purpose vehicles by firms like Apollo and Blackstone, are transferring leverage down the supply chain in ways that are difficult for the broader market to see .

Historical Parallels: Three Cycles That Broke

Man Group explicitly draws parallels to three prior sector-driven credit cycles where a wave of debt-fueled investment ended in dislocation. The pattern, they argue, is consistent: a compelling new sector attracts enormous debt capital, financial discipline erodes, and when demand or pricing assumptions prove overoptimistic, credit markets absorb heavy losses .

• Internet/Telecoms (Early 2000s): Massive overinvestment in fiber and telecom infrastructure led to a wave of defaults after the dot-com bubble burst .
• European Banks (2012): A sovereign debt crisis triggered bank bailouts and a credit market freeze .
• Energy (2015–2016): The oil price collapsed after a debt-fueled US shale drilling boom, producing a wave of high-yield defaults .

The current AI buildout, Man Group argues, shares these structural fingerprints. Each cycle looked different in its specifics but followed the same arc of excessive optimism, deteriorating lending standards, and painful repricing when reality failed to match expectations .

What Happens If the AI Credit Bubble Bursts?

The potential fallout extends beyond the tech sector. If AI companies cannot rapidly increase revenues to service their massive debt loads, the interconnected nature of modern financial markets means losses could spread through an interconnected set of financial institutions, as outlined by U.S. Senate Democrats in a January 2026 letter to the Financial Stability Oversight Council .

Oliver Wyman has projected $6 trillion in total AI capital spending between now and 2030, with an increasing share being debt-financed, much of it in off-balance-sheet vehicles remote from the cash-rich tech titans . If half of that is debt-financed, the credit buildup would exceed all broadband infrastructure investment since the beginning of the internet .

The trigger for a repricing, according to economists like Ruchir Sharma, could be higher interest rates, which would reduce the availability of cheap capital that has been fueling AI investment and put downward pressure on growth-stock valuations . The recent history of the leveraged loan market, which saw a sharp 34% decline in Q1 2026 activity compared to the previous year—particularly in software and services—suggests the credit market is already beginning to differentiate between winners and losers .