Blackstone president calls for more equity to fund AI buildout as debt risks mount

Blackstone President Calls for More Equity to Fund AI Buildout as Debt Risks Mount Jon Gray warns that the $5+ trillion AI infrastructure spending wave is pushing leveraged balance sheets to the edge — and equity investors must step up or face a repricing of the entire debt market. --- The race to build the AI infrastructure stack is generating a staggering amount of debt. Blackstone President Jon Gray is now publicly calling for a fundamental shift in how the capital markets fund AI buildout — arguing that debt investors are being asked to absorb concentration risk they are not structured to handle, and that private equity and institutional capital must provide the equity backbone that keeps leveraged financing sustainable. His warning arrives at precisely the moment when the numbers are becoming impossible to ignore.

The Scale of the AI Infrastructure Spending Wave

The data is unambiguous. Global capex directed toward AI infrastructure — data centers, semiconductor fabrication, networking equipment, and related compute infrastructure — has crossed well beyond $5 trillion cumulatively, with projections suggesting that figure could double or triple by 2030. That spending is overwhelmingly debt-financed. Hyperscalers like Microsoft, Google, Amazon, and Meta have issued billions in high-yield and leveraged debt to fund data center construction, GPU procurement, and network buildout. Private equity firms, sovereign wealth funds, and infrastructure funds have followed suit, deploying billions in leveraged buyouts and project finance to acquire or build AI-related assets. The problem, as Gray frames it, is one of concentration risk. When a small number of borrowers — the hyperscalers and the largest PE sponsors — account for a disproportionate share of all AI-related debt issuance, the entire credit market becomes fragile. A default or even a downgrade by one major player doesn't just hurt that borrower; it ripples through the supply chain of underwriters, rating agencies, and the broader leveraged loan and high-yield bond markets.

Why Debt Alone Cannot Finance the AI Buildout

Gray's argument is structural, not ideological. The capital structure of AI infrastructure projects is inherently complex — long-duration assets requiring massive upfront capex, uncertain technology pathways, and long payback periods. Debt markets, by their nature, optimize for shorter-duration, cash-flow-backed instruments. When you layer 20-year project finance debt on top of an asset whose revenue model depends on AI compute demand that is still evolving, you create a mismatch that amplifies risk. "The debt markets are being asked to do something they were never designed to do," Gray told CNBC. "We are seeing a concentration of risk in a very narrow set of borrowers, and the equity markets need to step in and provide the capital that absorbs that risk over the long duration." This is not a call for capital discipline in AI investment — it is a call for capital structure innovation. The distinction matters. Gray is not saying AI buildout should slow down. He is saying that the financing model needs to evolve to match the scale and duration of the opportunity.

The LDI Parallel: Lessons from the 2022Debt Crisis

There is a direct analogy that Gray and market observers are drawing to the 2022 gilt crisis, when the UK's gilt market experienced a disorderly repricing driven by a concentrated set of leveraged holders — pension funds and insurance companies — forced to sell into a falling market because their liability-matching models could not accommodate the speed of the decline. The result was a liquidity spiral that forced central bank intervention at the worst possible time. The AI debt concentration problem is structurally similar. A small group of leveraged investors — pension funds, insurance companies, sovereign wealth funds — are now major holders of AI-related debt. If those borrowers face margin calls, refinancing walls, or credit deterioration, the forced selling dynamics could produce the same kind of liquidity crisis, just in a different market.

What an Equity-Led AI Financing Model Looks Like

Gray's prescription is for a more diversified capital stack — one where equity providers absorb the first-loss position and provide the long-duration capital that debt markets structurally cannot. This would involve: Increased private equity and infrastructure fund participation in AI asset ownership, providing patient capital that does not need to be repaid on a short schedule. Public market equity listings for AI infrastructure assets, creating transparent, liquid markets where early investors can monetize positions without forcing asset sales. Regulatory clarity on how pension and insurance funds can hold concentrated positions in AI-related credit, reducing the forced-selling dynamic. The practical effect would be to dilute the concentration of debt risk across a broader base of capital providers, making the entire system more resilient to shocks.

The Market Response and What It Means for Investors

Not everyone is convinced. Some market participants argue that the hyperscalers have sufficient cash flows and balance sheets to service the debt they have already issued, and that the risk is overblown. Others note that the AI buildout is still in its early stages, and that the revenue realization timeline is longer than historical infrastructure projects. But the data on refinancing risk is already visible. A growing share of AI-related debt is maturing in the 2025–2028 window, precisely when interest rates remain elevated and the revenue visibility for some AI infrastructure assets is still uncertain. Refinancing walls are building — and they are building in a concentrated pool. For investors, the key takeaway is this: the AI infrastructure theme is real and massive, but the financing model is fragile in its current form. The next phase of alpha will come not just from identifying which AI bets will win, but from understanding which financing structures will survive the inevitable corrections. ---

Frequently Asked Questions (FAQ)

Q: Why is Jon Gray specifically calling for more equity funding rather than more debt? A: Because debt markets are structurally ill-suited to fund long-duration, high-capex AI infrastructure assets. Concentrated debt exposure creates systemic risk similar to what happened in the 2022 gilt crisis. Equity provides long-duration capital that can absorb first-loss risk without triggering forced selling in the debt markets. Q: How much of the AI infrastructure spending is currently debt-financed? A: The majority — well over $5 trillion cumulatively — is debt-financed, with hyperscalers, private equity firms, and infrastructure funds issuing billions in leveraged debt to fund data centers, compute infrastructure, and networking buildout. This concentration is the core risk factor Gray is warning about. Q: What happened in 2022 with the gilt crisis, and why does it matter for AI debt? A: In 2022, a concentrated group of leveraged holders in the UK gilt market was forced to sell into a falling market, creating a liquidity spiral that required emergency central bank intervention. Gray draws a direct parallel: AI-related debt concentration could produce the same dynamic in credit markets, just with different underlying assets.