Ask Finn← Discover
YOUR MONEY

Blackstone's Jon Gray Warns AI's Trillion-Dollar Buildout Is Drowning in Debt

By Hayden Walsh · Wednesday, November 4, 2026
Finn's Take· TL;DR
  • AI infrastructure relies heavily on debt, creating dangerous concentration risk across narrow set of lenders and companies needing urgent shift to equity financing.
  • Tech giants like Meta and Google expected to issue significant equity, while structured deals aim to reduce debt burden in trillion-dollar buildout.
  • Despite massive growth potential, current debt-heavy funding model caps lender returns while exposing them to enormous risk in uncertain sector.
See this from any side — with sources:
Left takeNeutralRight take

A Warning From the Top of Wall Street

One of the most powerful figures in global finance is raising a red flag about how the artificial intelligence boom is being funded — and his message is blunt: the industry has borrowed too much, and something has to change. On November 3, Blackstone President Jon Gray made the case that the AI infrastructure buildout urgently needs more equity financing, warning that lenders are dangerously overexposed to a narrow slice of the market.

Speaking at the Hong Kong Monetary Authority's Global Financial Leaders' Investment Summit, Gray said the scale of capital required for AI is creating concentration risks for debt investors, both across the sector and at individual companies. In plain terms: too many lenders are betting too heavily on too few companies — and if something goes wrong, the fallout could be severe.

The Numbers Are Staggering

PIMCO CEO Emmanuel Roman cited a $10 trillion five-year spending estimate and $550 billion in AI debt issued so far in 2026 as evidence that the financing machine is running at full throttle — perhaps too full. Roman compared the scale with a global corporate bond market of about $20 trillion and warned that AI financing could eventually become excessive.

Gray said lenders receive interest and repayment of principal, but do not share in a company's growth. That's the core problem with a debt-heavy model: lenders take on enormous risk but are capped in what they can earn. Equity investors, by contrast, can benefit if the companies they back become enormously valuable — making equity a more natural fit for a sector where the upside potential is massive but so is the uncertainty.

Who Would Raise the Equity — and How

Gray said major tech firms including Meta, Google, and Oracle are expected to make significant equity raises. That means more shares coming to market from some of the world's most valuable companies. Gray said financing the buildout would require a combination of public markets, banks and private capital, with companies increasingly turning to structured transactions to limit the amount of debt they take on.

Gray pointed out that Blackstone reached an agreement last week with Bank of America, Citi and Morgan Stanley to secure $63 billion in financing for purchasing Broadcom's networking hardware and AI chips intended for Anthropic. The deal underscores just how massive individual transactions in this space have become — and why the broader financing structure deserves serious scrutiny.

Bullish on AI, Cautious on How It's Funded

Despite those financing challenges, Gray remained bullish on AI infrastructure, saying demand for computing power was "running well ahead of supply." He said Blackstone portfolio companies' spending on AI models grew 20-fold over the past year. That kind of explosive growth is exactly why the funding conversation matters so much — the numbers involved are no longer abstract.

Hyperscaler capex at about 2.5% of US GDP means the AI buildout is no longer a niche corporate story. It's a systemic one. As concentration risks mount and debt investors get skittish, the industry faces a capital structure reckoning. Gray's message isn't that AI is in trouble — it's that the way the world is paying for it needs to grow up fast, before the debt markets force a harder correction on their own terms.

Have a question about this story?
Ask Finn — answers grounded in this article, from any viewpoint.