Critical Warning in AI Bubble: Profits Stem from Investors, Not Customers

The global financial markets' most discussed artificial intelligence (AI) boom has been shaken by a stark analysis from Apollo Global Management Chief Economist Torsten Slok. Slok argues that while the current AI rise appears to be a technological triumph, it is fundamentally an unsustainable financing bubble, with the industry's most profitable sector kept afloat by investor funds from loss-making segments.

The Inversion of the Value Chain: Chipmakers Profit While Models Bleed

Slok's analysis, utilizing data from PitchBook and Bloomberg, exposes the paradoxical structure of the AI value chain. Contrary to traditional economic models, companies selling the end product are recording the biggest losses, while upstream suppliers secure record profits.

■Silicon and Equipment (Chipmakers like Nvidia, AMD): This segment, holding the highest profit margins in the AI value chain, boasts an operating margin of 41%.

■Models and Applications (OpenAI, Anthropic, etc.): Companies providing the final product to consumers are seeing operating margins deep in the red at -59%.

■The Apollo Chief Economist asserts that this disparity proves "AI boom profits are currently being funded by investors rather than earned from customers."

The Trillion-Dollar Gamble: Divergence in Debt and Return Calculations

Despite Goldman Sachs projecting AI investments to swell beyond $1 trillion by 2026, Slok and other experts note that these expenditures have not yet translated into economic productivity. Data from the Bank for International Settlements (BIS) and Bank of America highlights a concerning surge in debt accumulation by tech giants to fund these endeavors.

■Major hyperscalers raised $121 billion in debt in 2025, four times the average annual issuance over the previous five years.

■Citing Oracle as an example, the report notes the company has a negative cash flow of $23.7 billion, $130 billion in outstanding debt, and $260 billion in lease commitments for projects yet to begin.

■If giants like Microsoft, Google, Amazon, and Meta decide to halt their quarterly spending of over $30 billion on GPUs and data centers, the assumed "permanent supercycle" faces an immediate threat of collapse.

Markets might read this merely as a tech sector crisis, but as a maritime economist, I see the tsunami this will create for the global supply chain. A sudden halt in semiconductor demand could cause a collapse in cargo traffic from Asia to the West, while energy and grid investments could be shelved. A constriction in AI financing would act as a severe brake on freight markets and energy commodity supplies, potentially flipping current inflationary pressures into a deflationary shock.