Global technology stocks slid on Monday as a crescendo of warnings from inside the AI industry itself sparked a broad reassessment of the sector’s growth assumptions. The selloff swept from Seoul to Silicon Valley, hitting chipmakers, cloud giants and the power-equipment suppliers that have ridden the data-center buildout.
The moves across markets
In Asia, SK Hynix dropped more than 6 percent and Samsung Electronics shed over 4 percent. SoftBank, a major OpenAI backer, fell 10 percent in Tokyo. European names followed: ASML declined more than 5 percent, Infineon lost over 7 percent, and Nokia slipped about 8 percent. Siemens Energy and Schneider Electric, both tied to data-center expansion, also traded lower. U.S. premarket action mirrored the tone. Micron fell roughly 5 percent, Intel nearly 6 percent, and Nvidia was down almost 3 percent. Microsoft, Amazon and Alphabet edged lower.
The catalyst was internal
The trigger was not a macro data point but a cluster of safety resignations and public statements last week. Jacob Coxon, an Anthropic researcher who previously worked at OpenAI, quit with a warning that both labs were “gambling with our lives.” His colleague Evan Hubinger put the probability of an AI-driven human extinction event within a decade at more than 10 percent. The comments ignited a debate that reached Anthropic chief executive Dario Amodei, who published an essay on Saturday calling for slower development of advanced systems.
Valuation exposure is the issue
The market reaction reflects a simple structural fact: much of the expected return from AI-exposed equities is already priced into future earnings growth. If that growth path comes under regulatory or reputational threat, the repricing can be violent. The selloff is not about current revenue; it is about the durability of the multiple assigned to that revenue.
What to watch next
The next signal will come from whether the safety debate translates into concrete policy, export controls, licensing regimes, or mandatory audits, that could slow capital deployment. Until then, the trade remains a bet on whether the industry can self-regulate fast enough to keep the buildout on schedule.