Anthropic Chief Executive Officer Dario Amodei spearheaded a coalition of prominent AI industry figures calling for a deceleration in AI capability advancement, prompting a broad decline in artificial intelligence-related stocks on Monday. SoftBank, the primary investor in OpenAI, experienced a substantial drop in its share price, while global chip manufacturers including SK Hynix, ASML, Micron, and Intel all moved lower in tandem.

The selloff emerged against a backdrop of intensifying debate over the risks associated with rapid AI model performance iteration, with discussions reaching a crescendo last week. On February 16, 2026, in Bangalore, India, Anthropic hosted its developer summit, where co-founder and CEO Dario Amodei appeared as a featured speaker. Following Amodei's call to moderate the pace of AI capability research and with other technology executives voicing support for the proposal, AI-related equities across global markets retreated on Monday.

Investors harbor concerns that an industry-wide slowdown in AI research and development could produce cascading effects across the supply chain, curtailing the scale of technology deployment and real-world application. In Asian markets, South Korean bellwethers SK Hynix and Samsung Electronics closed sharply lower, declining more than 6% and 4% respectively. SoftBank, a major backer of OpenAI, plunged 10% in Japanese trading. European markets mirrored the downturn during early session trading, with semiconductor and other AI-linked stocks suffering heavy losses. Chip equipment giant ASML fell over 4%, Nokia dropped approximately 5%, and Infineon declined more than 6%. Companies across the data center construction supply chain also saw their shares retreat.

Pre-market trading in the United States reflected similar weakness among major AI names: memory chip maker Micron fell roughly 5%, Intel dropped nearly 6%, and NVIDIA declined more than 2%. The catalyst for this round of selling stems from heated debate over the risks of rapid AI model iteration, with that wave of public discourse reaching its zenith last week. Jacob Caukson, an Anthropic researcher who previously worked at OpenAI, publicly stated that Anthropic and OpenAI are "gambling with humanity's fate." Following those remarks, Evan Hubinger, an Anthropic safety researcher, voiced his assessment that there exists a greater than 10% probability AI could "lead to the extinction of all humanity" within the next decade.

The related social media posts ignited a firestorm of discussion, prompting senior executives across the AI industry to collectively weigh in. Dario Amodei published a blog post titled "We Must Control the Pace of Frontier AI Development," urging a slowdown in the rate of AI capability iteration. Amodei stated: "We must decelerate the pace at which AI model capabilities evolve."

Amodei's essay galvanized a rare consensus among executives at rival AI companies. OpenAI CEO Sam Altman responded on Saturday, endorsing Amodei's perspective and asserting that AI enterprises need to "manage the pace of frontier technology advancement" rather than halting research entirely. SpaceX CEO Elon Musk, who has warned about AI's potential dangers for years, posted on social platform X: "Dario is right."

Market participants worry that any slowdown in AI development could disrupt chip procurement and computing power acquisition, sectors currently absorbing hundreds of billions of dollars in capital expenditure. Zoe Gillespie, senior director at RBC's Bruin Dolphin unit, remarked on Monday during the "European Squawk Box" program: "The current market rally is built substantially on expectations of AI-driven growth and productivity gains... If that growth narrative wobbles, subsequent equity market performance could come under pressure."

"A significant portion of current earnings expectations already factors in future profit growth for these companies. Should that earnings growth be threatened, markets could enter a period of turbulence."

Notably, Amodei advocates for decelerating frontier AI capability iteration rather than imposing a complete moratorium on research, explicitly noting that "technological progress will still maintain a relatively rapid pace." Altman posted on social platform X on Monday that "managing pace does not mean stopping research." "Technological development is already fast and will continue to evolve quickly going forward, but the overall tempo should be slower than an unconstrained state; interventions such as safety assessments and risk monitoring inherently carry significant costs."

Ben Ballinger, global technology research head at Quilter Cheviot, told that "the pace of development may moderate somewhat, but the magnitude of industry transformation remains enormous." "Even if model training and deployment timelines are slowed, a massive inference computing gap persists within the industry, with demand far outstripping supply. Therefore, even a modest cooling in development pace may not necessarily inflict damage on corporate revenues."