In a blog post published on Friday, Apollo Chief Economist Torsten Slok laid bare a striking structural imbalance in the artificial intelligence value chain: the segments with the highest profit margins—chipmakers and equipment suppliers—are sustained by capital raised from investors rather than revenue earned from end customers. The finding challenges the conventional business logic in which companies selling finished products to consumers command the strongest margins.

Slok categorized AI companies into four groups: models and applications, cloud and compute, energy and grid, and silicon and equipment. Drawing on data from Pitchbook and Bloomberg for companies including OpenAI, Anthropic, Microsoft, Amazon, Constellation Energy, Nvidia, AMD, and Micron, he calculated that silicon and equipment firms—such as chipmakers—hold the highest profit margin in the AI value chain at 41%. Models and applications companies—the layer closest to end users, where ChatGPT, Claude, and other consumer-facing products compete—post a -59% operating margin. The gap underscores a reality that the AI industry's headline revenue figures often obscure: the money flowing downstream to semiconductor makers originates largely from venture funding, debt issuance, and corporate capex budgets rather than paying customers.

This sharp divergence, Slok cautioned, reflects the fact that money flowing into the AI boom stems not from organic demand for AI applications but from investors betting on the next technological revolution.

"AI boom's profits are currently being funded by investors rather than earned from customers," Slok wrote in his Apollo analysis. "The upstream margins are real, but they are paid for out of capital raised by the layer losing money, not out of cash generated by end demand."

Goldman Sachs now projects AI investments to swell beyond $1 trillion in 2026—a sum rivaling the total U.S. nonresidential fixed investment in information processing equipment and software in a typical year. Thus far, however, the technology has produced no significant changes in economic productivity or profit margin growth outside the Magnificent Seven—Apple, Microsoft, Nvidia, Amazon, Alphabet, Meta, and Tesla—which together account for roughly a third of the S&P 500's market capitalization. Should AI financing decelerate, Slok warned, the lopsided margin structure could destabilize the entire industry.

"The bottom line is that the most profitable part of the AI value chain depends on the least profitable part continuing to grow revenue or raise capital," he concluded. "Capital can bridge the gap for a while, but not indefinitely. And therein lies the risk: will the ROI show up for AI's end customers fast enough to sustain the spending that is generating those upstream margins?"

Broader Concerns About Unsustainable AI Expansion

Slok is not alone in flagging AI's heavy dependence on investment capital. In its annual report published in June, the Bank for International Settlements—the Basel-based institution that serves as a central bank for central banks—observed that AI spending—driven primarily by the five major hyperscalers—is outpacing earnings and free cash flow, prompting these companies to issue debt to raise additional financing. A Bank of America analysis from last November found that in 2025, those five hyperscalers issued $121 billion in debt, four times the average annual debt levels these firms issued over the previous five years.

"Disappointment in returns could trigger a sudden pullback in financing and turn the capex boom into a protracted investment bust, with potential knock-on effects on financial conditions," the BIS said in its report. "Should hyperscalers slow or halt the aggressive pace of capex deployment, many borrowers across the supply chain could struggle to replace lost revenue and service their debt."

Tech commentator Ed Zitron extended this argument further, contending that AI spending is even more precarious than it appears on the surface. In a Substack post published in June, he pointed to Oracle as a case study: the company posted a negative cash flow of $23.7 billion as of the end of fiscal 2026, with nearly $130 billion in outstanding debt and $260 billion in lease commitments for AI infrastructure projects that have yet to commence. Oracle's massive AI buildout is in service of OpenAI, with which it signed a $300 billion deal last September.

"Oracle's existence—and Larry Ellison's personal wealth—hinges on whether OpenAI can make good on its promise to spend $300bn in compute," Zitron wrote.

He described this dynamic as the "most-obvious and under-discussed part of the AI bubble." While 13-figure hyperscaler capital expenditures are fueling a semiconductor boom, there is little evidence so far that the current AI boom will translate into widespread applications of technology sufficient to justify all this spending.

More alarming than Oracle's potential failure to deliver on its $300 billion commitment, Zitron argued, is the prospect of other major tech companies pulling back on their own exorbitant AI outlays.

"If Microsoft, Google, Amazon and Meta decide that it's time to stop spending $30 billion or more a quarter on GPUs, RAM, storage, and data center construction," he said, "that'll tear a hole in the side of what people assume is a permanent supercycle."