Apollo CEO Warns AI Buildout Needs $8 Trillion
Marc Rowan says the AI infrastructure boom could require as much as $8 trillion in financing, with chip-backed debt emerging as a key tool.
Updated on Nov. 17, 2026 in Data Centers

Apollo Global Management CEO Marc Rowan warned on CNBC's "Squawk Box" that the AI infrastructure boom could require up to $8 trillion in financing. Apollo is participating in a roughly $40 billion financing arrangement involving Nvidia chips for SpaceX.
Why it matters
The scale of capital needed for AI computing infrastructure could strain lenders and push companies to rely on structured financing backed by equipment. Concentrated exposure to a small group of AI-driven companies raises questions about how risk is priced and distributed across credit markets.
Chip-backed financing can give lenders security interests in equipment with repayment schedules reaching zero within five years, while neocloud providers carry debt trading near 15% yields. It remains unclear how these structures will perform if AI demand or chip values shift.
The players
Apollo Global Management
Apollo Global Management is a major alternative asset manager that participates in large-scale credit and infrastructure financing.
Marc Rowan
Marc Rowan is the chief executive of Apollo Global Management and a prominent voice on private credit and financing markets.
The details
Rowan said the AI, data center, energy and infrastructure spending cycle is exceeding investments from previous technological revolutions over the past 200 years. He noted that borrowers must offer higher returns and greater downside protection to lenders taking on concentrated AI exposure, and that equity investors can benefit from unlimited upside while credit investors are generally limited to interest and principal.
Timeline
Over the past 200 years, the AI spending cycle has exceeded investment levels of previous technological revolutions.
Over the next several years, hyperscalers could become six of the 10 largest investment-grade debt issuers.
Over five years, chip-backed financing repayment schedules can bring balances down to zero.
Roadmap
Rowan's comparison places the AI buildout against roughly 200 years of technological revolutions, suggesting capital needs now exceed those earlier cycles. The concentration of spending among a small group of companies makes the financing question central to the sector's trajectory.
Higher financing costs for AI infrastructure could eventually shape the pricing and availability of cloud computing services used by businesses and consumers. Investors in credit markets may also see exposure to concentrated AI borrowers through funds and retirement holdings.
The takeaway
The AI infrastructure boom is becoming as much a financing story as a technology story, and the terms lenders demand will shape which companies can keep building. Readers tracking tech investment should watch how credit markets price chip-backed debt.
Further reading
For the infrastructure side of this buildout, see our Data Centers coverage.






