AI Infrastructure Debt Could Trigger Massive Private Credit Bubble
Chamath Palihapitiya warns that massive debt used to fund AI infrastructure, including chips and data centres, could lead to a significant financial bubble.
The relentless race to build out artificial intelligence infrastructure is increasingly being fueled by massive amounts of debt, raising serious concerns about a potential financial bubble. Chamath Palihapitiya, the CEO of 8090 and Social Capital, has sounded the alarm, questioning whether the heavy reliance on private credit to fund the AI revolution could eventually lead to significant financial instability.
As technology giants pour billions into chips, data centres, and other essential hardware, the role of private credit has expanded dramatically. Palihapitiya noted that the AI buildout is being funded by debt at a much higher rate than just a year ago, prompting his firm to investigate whether this trend is creating a precarious bubble.
Record-Breaking Deals and Massive Spending
The scale of these financial arrangements is unprecedented. Palihapitiya highlighted a massive USD 35 billion vehicle led by Apollo and Blackstone in the second quarter of 2026, specifically designed to purchase chips and lease them to Anthropic. This transaction was described by UBS as the largest private credit deal on record.
To put the scale into perspective, Palihapitiya pointed out that direct lenders made approximately USD 23 billion in loans to private-equity-backed U.S. companies during that same quarter—meaning the single AI chip deal far outpaced general private equity lending.
The financial requirements for the AI era are set to skyrocket. Industry forecasts suggest a staggering trajectory for capital expenditure among tech leaders:
- 2026: Alphabet, Amazon, Meta, Microsoft, Oracle, and SpaceX (xAI) are expected to spend around USD 870 billion.
- 2027: Spending is projected to surge to more than USD 1.3 trillion.
These astronomical costs are being met through a mix of debt, leases, equity issuance, and various other financing methods.
The Growing Debt Trap
The expansion of direct lending in the United States has been explosive. Palihapitiya noted that direct loans to U.S. borrowers expanded nearly 25 times, jumping from USD 96 billion in 2010 to USD 2.4 trillion by 2025. Even more striking is the lending to software and technology companies, which grew approximately 45 times, rising from USD 22 billion to over USD 1 trillion in the same period.
This rapid growth raises a critical question: Are lenders adequately pricing the specific risks associated with AI?
Data from the Bank for International Settlements (BIS) suggests a potential mismatch. In January, the BIS found that private loans to AI-related companies carried an average spread of 6.2 percentage points above the base rate, which is only marginally higher than the 6.1 percentage points seen in other private credit loans. Palihapitiya suggested this narrow gap indicates a dangerous trend: "Lenders are underestimating AI-related risks, equity investors are overestimating future profits, or both."
A Double-Edged Risk for the Tech Sector
The potential fallout from an AI-driven credit bubble could hit the industry from two different directions:
- Disruption of Existing Software: AI could disrupt established software companies that have already taken on significant debt through private credit markets. If AI renders their products obsolete, their revenues could plummet, leaving them unable to repay their loans.
- Repayment Pressures on AI Leaders: Conversely, the companies investing heavily in AI infrastructure may face severe repayment pressures if the expected returns on their massive investments take longer to materialize than anticipated.
As the market begins to distinguish between different types of AI financing deals through varying trading prices, Social Capital is conducting research into refinancing risks and investor withdrawals. The goal is to determine if the current trajectory of AI investment will turn private credit into a major source of global financial stress.