Global artificial intelligence stocks came under heavy pressure on Thursday after Broadcom chief executive Hock Tan warned that parts of the market were behaving less like disciplined investors and more like casino gamblers. The comments landed at a sensitive moment for the sector, where share prices have been propelled by extraordinary expectations for AI adoption, infrastructure spending and future monetisation, leaving little room for disappointment.
Tan's remarks sharpened an already growing debate on whether the AI boom is being priced on durable earnings power or on speculative enthusiasm. His warning that investors are treating AI "like gambling" resonated across trading desks because it struck at the core of the current market narrative: that massive capital expenditure on chips, data centres and model training will inevitably translate into outsized profits. The selloff that followed suggested that many investors are increasingly uneasy about how much of that future has already been discounted into valuations.
Valuation Anxiety Builds
The immediate market reaction was broad and swift. Shares linked to the AI trade weakened across major exchanges, with South Korea's Kospi and Japan's Nikkei both declining as investors reassessed exposure to semiconductor and technology names. The move reflected a wider global de-risking in which traders trimmed positions in companies most closely associated with the AI buildout.
The pressure was not limited to hardware suppliers. Tan's comments also revived scrutiny of the software and platform layer, particularly companies that have become symbols of the AI revolution. He questioned whether OpenAI, despite its strong revenue growth, can ultimately achieve sustainable profitability. That concern matters because the AI ecosystem is still heavily dependent on continued funding, rapid user adoption and a clear path to monetisation. Revenue growth alone, analysts say, does not guarantee that the economics of model development, inference and infrastructure can support the valuations now attached to the sector.
For investors, the issue is not whether AI is transformative. Few doubt that it will reshape enterprise software, cloud computing, chip demand and productivity. The question is whether the market has moved too far, too fast, and whether the spending required to maintain leadership in AI can be justified by future cash flows. Tan's comments effectively put a spotlight on that gap between technological promise and financial reality.
Asia Feels The Shock
The selloff rippled through Asian markets, where technology and semiconductor shares have been among the strongest performers in the global AI rally. South Korea and Japan are especially exposed because of their deep links to memory chips, advanced manufacturing and electronics supply chains. As sentiment turned, investors moved to lock in gains and reduce risk in names that had benefited from the surge in AI-related demand expectations.
India, however, saw a different reaction. Indian IT stocks rallied as global technology shares weakened, reflecting a rotation that often benefits service providers when investors become more cautious on capital-intensive AI plays. The move suggested that the market is beginning to distinguish between firms that sell AI infrastructure and those that may profit from enterprise adoption, consulting demand and software integration without bearing the same level of upfront investment risk.
That divergence is important for Indian markets. While domestic IT companies have their own growth challenges, they are often viewed as relatively insulated from the most extreme valuation swings in the global AI hardware trade. A global pullback in expensive tech names can, at least temporarily, improve the relative appeal of Indian software exporters.
Profitability Question Returns
Tan's remarks also arrive at a time when investors are increasingly asking how long AI spending can outrun earnings. The sector has been supported by a powerful cycle of capital expenditure, with cloud providers, chipmakers and model developers all racing to expand capacity. But the scale of that investment has raised concerns that returns may take longer to materialise than markets currently assume.
The warning from a senior industry executive carries weight because it comes from inside the ecosystem rather than from a distant critic. It suggests that even among beneficiaries of the AI boom, there is recognition that sentiment may be overheating. For markets, that is often the point at which a narrative-driven rally becomes vulnerable: when the same companies that helped build the boom begin to question its pricing.
For now, the selloff appears to be a reassessment rather than a rejection of AI itself. But the message from the market was clear. Investors are no longer willing to assume that every dollar spent on artificial intelligence will produce immediate or even predictable returns. In a sector where expectations have become as important as earnings, that shift alone can move billions of dollars in market value within hours.