Asian Technology Sell-Off Deepens as Investors Question the Cost of the AI Boom

LONDON — Technology shares across major Asian markets fell sharply as investors reassessed the high valuations of semiconductor manufacturers and other companies closely connected with the artificial-intelligence boom.

South Korea's Kospi suffered particularly heavy losses, while Japan's Nikkei and shares in major Taiwanese chip companies also declined. The sell-off reflected concern that expectations surrounding artificial intelligence had risen faster than the earnings and cash flows currently being produced by many companies.

The market pressure does not mean that artificial intelligence has stopped growing or that the technology has failed. It shows instead that investors are becoming more selective about which companies can convert extensive spending on data centres, advanced chips, cloud services and power infrastructure into dependable revenue.

That distinction became clearer after contrasting results from Microsoft and Meta. Microsoft's latest performance strengthened confidence that demand for its cloud and AI products was supporting revenue growth, while Meta's shares fell after the company reported heavy spending and a sharp decline in free cash flow.

Meta reported second-quarter revenue of approximately $60.8 billion, representing strong year-on-year growth. However, free cash flow fell substantially as investment in artificial-intelligence infrastructure accelerated, and the company raised its expected capital-expenditure range for 2026.

Investors are increasingly asking three questions: how quickly AI services will generate direct revenue, whether companies can maintain profit margins while infrastructure costs rise, and whether new competitors will reduce the pricing power of established chip and cloud providers.

Asian semiconductor companies are especially exposed because many experienced steep share-price increases during the earlier phase of the AI investment cycle. When valuations are already high, even record profits may disappoint investors if earnings fall below the most optimistic expectations.

The decline was not uniform across global markets. London's FTSE 100 proved more resilient because it has greater exposure to banks, energy companies, miners and consumer businesses than to the large technology groups dominating some Asian and American indices.

The current correction should therefore be described as a reassessment of prices and expectations, rather than definitive evidence that the wider AI industry is collapsing.

Why it matters

Pension funds, individual investors and technology businesses have significant exposure to the AI investment cycle. A prolonged correction could affect corporate fundraising, semiconductor orders and plans for new data centres, while stronger-than-expected earnings could quickly restore confidence.