Shares in Samsung Electronics and TSMC fell as investors began questioning whether the AI chip boom rests on borrowed money, Capital Brief reports. The sell-off hits the two companies that manufacture the overwhelming majority of the world's advanced logic chips, and it signals a shift in how the market prices AI infrastructure risk.

Why are the foundries sliding?

The trigger is not a capacity problem or a technology stumble. It is a financing question. Investors are increasingly focused on how much of the demand for AI accelerators and high-end silicon comes from customers funding purchases with debt rather than cash flow from deployed systems.

Samsung and TSMC sit at the top of that supply chain. Their order books reflect hyperscaler capital spending on AI data centers, and their shares have traded as a proxy for the sector's health. When the market starts to discount the durability of AI-related capital expenditure, the foundries feel it directly.

"Samsung, TSMC shares fall as investors question debt-fuelled AI chip boom," Capital Brief reported, capturing a session in which both stocks declined.

What does debt-financed demand change?

The distinction matters commercially. Demand underwritten by cash-generating services behaves differently from demand underwritten by leverage. If AI infrastructure buyers are borrowing to fund chip purchases, then rising interest costs, tighter credit conditions or slower AI revenue monetization could translate into order delays or cancellations upstream at the foundries.

That chain runs in one direction only. A hyperscaler that defers a data center build pushes the pause button on wafer orders months before end-market demand actually proves or disproves the AI thesis. Investors in Samsung and TSMC are therefore pricing in credit conditions as much as semiconductor fundamentals.

The concern lands on two companies with very different competitive positions:

• TSMC manufactures the most advanced AI accelerators and commands the leading share of leading-edge foundry output, making it the most direct beneficiary — and the most exposed — of any slowdown in AI chip spending.
• Samsung competes at the advanced node level while also carrying a memory business, giving it a second lever on AI demand but a second point of exposure if that demand proves debt-dependent.

Is this a correction or a trend?

A single down session does not settle the question. What it does establish is that the debt-financed character of parts of the AI build-out is now an explicit variable in how investors value the sector's suppliers, rather than background noise.

Until recently, foundry shares traded on capacity tightness, node leadership and pricing power. Capital Brief's report marks the arrival of a fourth factor: the balance-sheet health of the customers placing the orders.

For Samsung and TSMC, the coming quarters will test whether AI-driven wafer demand holds up as credit scrutiny deepens — and whether investor confidence in the chip cycle can survive a closer look at who is borrowing to pay for it.