Taiwan's central bank governor has raised a stark warning: the artificial intelligence boom driving record stock gains is built on a foundation of dangerous financial excess. Governor Yang Chin-long told lawmakers Thursday that while AI growth is real, the way companies are financing it isn't.
"We do have concerns about the possibility of an AI bubble," Yang said at a parliamentary hearing. "AI is driven by real growth potential, but it's the possibility of over-expansion via over-leveraging that concerns us." The distinction matters enormously. It's the difference between sustainable prosperity and a financial reckoning that could devastate Taiwan's economy and ripple globally.
The central bank's board acknowledged these risks last month when it decided to hold interest rates steady, though the vote wasn't unanimous. Yang explained the decision by pointing to a troubling economic imbalance: while the tech sector surges, traditional industries continue to underperform. That's not a sign of health. It's a sign of an economy becoming dangerously concentrated.
The Leverage Problem
What worries Yang most is the mechanism driving the boom. Tech companies aren't just investing in AI because demand justifies it—they're borrowing aggressively to finance capital expenditures they hope will pay off. This is speculative investment dressed up as strategic planning. When companies across an entire sector leverage themselves this heavily on the same bet, the system becomes fragile.
Taiwan's role in global AI infrastructure makes this concern urgent. The island is the linchpin of the world's semiconductor supply chain. Taiwan Semiconductor Manufacturing Co., or TSMC, manufactures the chips powering AI systems for Nvidia, Apple, and countless others. TSMC's dominance has sent Taiwanese stocks to record highs this year, enriching shareholders while the underlying financial structure grows more precarious.
Even TSMC itself is grappling with the cost side of this equation. Last month, the company acknowledged that while customer demand remains strong, rising component costs are eating into margins. That's a warning sign buried in an otherwise bullish statement. Customers are still optimistic about AI's future, but the economics are tightening.
Global Stakes
Nvidia CEO Jensen Huang's frequent high-profile visits to Taiwan—including a major trip in June for events like Computex and Nvidia GTC Taipei—underscore just how central Taiwan has become to the global AI supply chain. The island's prosperity is now intertwined with the health of this sector. But prosperity built on overleveraged corporate borrowing is prosperity on borrowed time.
Yang's warning reflects the tension between two competing imperatives: letting the AI boom run its course to maximize short-term growth, or tightening financial conditions now to prevent a catastrophic bust later. The central bank chose caution by holding rates steady rather than raising them. But Yang's public statements suggest that patience is wearing thin. The bank is watching. The risks are real. And Taiwan's policymakers know they can't afford to be caught flat-footed if the bubble deflates.
Why This Matters:
Taiwan's AI sector boom has created enormous wealth concentration among tech companies and their investors, but it's being financed through aggressive corporate borrowing that leaves the entire system vulnerable to sudden shifts in investor sentiment or market conditions. When financial regulators—even those in a country benefiting enormously from the boom—start publicly warning about bubbles and overleveraging, it's a signal that the current growth model is unsustainable. The risk isn't just to Taiwan's economy; it's to global supply chains that depend on Taiwan's chip industry. If companies across the sector are simultaneously overleveraged and facing rising costs, a correction could disrupt AI development worldwide. More fundamentally, this reveals how modern economies can create spectacular wealth gains while concentrating risk in ways that ordinary workers and traditional industries bear the cost of—whether through eventual recession, job losses, or the widening inequality between booming and declining sectors.