AI’s debt-backed expansion is testing investors’ confidence

The exposure runs through the supply chain

Alphabet’s plunge is a warning sign, not proof that the AI trade is breaking. But the industry’s financial structure means a pullback could travel quickly: many companies depend on circular financing, and a substantial share of the sector’s funding is debt-backed.

Data-center operators are among the businesses under pressure. CoreWave, an Nvidia customer, has seen its shares fall for months. Analysts call it a “canary in the coal mine” for AI-related stocks — an unsettling signal, though not a verdict on the industry.

Meanwhile, data-center expansion continues despite broad, bipartisan criticism. The scale of investment makes the downside harder to contain if demand weakens or the economy slows.

South Korea is doubling down

The concern is not limited to the United States. The Bank of Korea recently warned of a “semiconductor version of Dutch disease”: a fast-growing sector crowding out other export-oriented industries.

Yet South Korea is still increasing its exposure. President Lee Jae Myung promised today to invest boldly in AI and new sources of economic growth. Spending on AI and chips is set to double next year.

I think the tension is sharper than a simple question of whether AI demand will hold up. The same investment that may create a new growth engine could also make other parts of the economy more dependent on one sector.

The announcement leaves a key issue unresolved: how much of this expansion can be supported by durable demand rather than financing that relies on the sector’s continued growth? If that support falters, the risk is not just a repricing of AI stocks. It is a chain of exposure that reaches operators, investors and the wider economy.