South Korea’s Kospi index fell 4.3%, while Japan’s Nikkei declined 1.4%, according to market reporting published on Tuesday. Major regional technology companies came under pressure, with SK Hynix losing 5.3% and Samsung Electronics falling 3.8%. Kioxia Holdings in Japan also declined, while Taiwan Semiconductor Manufacturing Company recorded a smaller loss.
The weakness followed a turbulent period for AI-related equities, which had been among the principal beneficiaries of rising investment in computing capacity, advanced chips and data centres. Investors are increasingly questioning whether projected revenues will justify the substantial capital commitments required to develop and operate the infrastructure supporting AI services.
Concerns extend beyond chipmakers. Data-centre operators and cloud-computing businesses must finance expensive equipment and facilities while managing electricity costs, depreciation and the risk that computing hardware becomes outdated. Higher borrowing costs and tighter credit conditions could make these investments more difficult to sustain, particularly for companies carrying significant debt.
The market reaction also reflects uncertainty over the timing of returns. Companies can spend heavily on infrastructure before new products and services generate sufficient revenue. If demand grows more slowly than expected, investors may reassess valuations, financing plans and the expected useful life of computing assets.
Attention is now turning to earnings from major technology businesses, including Meta and Tesla, alongside reports from other companies exposed to the AI investment cycle. Results and forward guidance may help investors judge whether spending is translating into stronger revenues, improved productivity or durable competitive advantages.
The latest losses do not, by themselves, establish that demand for AI technology is weakening permanently. They do, however, show that investors are becoming more selective about which businesses can convert technological adoption into sustainable earnings.
For semiconductor manufacturers and infrastructure providers, the next phase of the AI cycle may depend less on announcing larger investment programmes and more on demonstrating financial discipline, credible demand and measurable returns.
