Blackstone President Jon Gray said this week that the artificial-intelligence buildout will require far more equity financing relative to debt, warning that lenders are taking on growing concentration risk as a handful of companies absorb an ever-larger share of AI infrastructure spending.
"There's going to be more equity needed across this space to keep up with these enormous demands," Gray said Tuesday at the Hong Kong Monetary Authority's Global Financial Leaders' Investment Summit, days before Thursday's broad selloff in AI and chip stocks.
Gray put a number on the scale of the buildout: the five largest hyperscale cloud companies are expected to spend a combined $820 billion on capital projects this year, about 2.5% of US gross domestic product, up from $415 billion in 2025. PGIM chief executive Emmanuel Roman, speaking at the same event, estimated AI-related capital spending could reach roughly $10 trillion over the next five years, with about $550 billion in AI-linked debt already issued in 2026 alone. Separately, Goldman Sachs estimated this month that data-center capital spending could total $8.5 trillion over five years.
Blackstone itself has significant exposure to the sector: the firm is a lender in the roughly $63 billion financing package announced in late October tied to Broadcom equipment leased to Anthropic, taking a second-lien position. Gray argued that as loans to AI infrastructure projects grow larger and more concentrated among fewer borrowers, lenders take on more downside risk for capped returns, while equity investors can absorb losses but also capture unlimited upside — a profile he said is better suited to projects whose payoffs remain uncertain.
He pointed to Meta, Google and Oracle as companies likely to raise significant new equity to fund their AI ambitions, alongside anticipated initial public offerings from SpaceX and Anthropic, as part of a broader shift away from debt-funded expansion.
The comments came as some of Blackstone's AI-adjacent credit investments have shown strain: bonds tied to some of the firm's data-center projects traded below 88 cents on the dollar at points during the year, even as the firm's AI investment chief, Jas Khaira, has argued that fears of a broader credit blowup are overblown because most data-center debt is backed by well-capitalized hyperscale tenants rather than speculative borrowers.