Top AI CEOs Call for Development Slowdown, Triggering Global Tech Stock Sell-off
Chief executives from leading AI firms including Anthropic, OpenAI, and xAI have publicly advocated for slowing down AI model iterations due to safety risks and potential threats to humanity. This collective warning triggered a sharp global sell-off across semiconductor and AI-related stocks in US, Asian, and European markets. This rare consensus among major AI leaders highlights escalating safety concerns as AI agents become capable of autonomous, potentially destructive actions on the internet. It has immediately rattled investor confidence, exposing the vulnerability of tech valuations that rely heavily on uninterrupted, rapid AI advances and massive capital expenditure. Anthropic CEO Dario Amodei warned that AI agents could gain internet-takeover capabilities within 6 to 12 months, causing hundreds of billions of dollars in losses, while OpenAI CEO Sam Altman announced OpenAI will pause IPO plans this year over safety concerns. Meanwhile, prominent investor Michael Burry dismissed the safety warnings as hype meant to cover up a slowing growth rate in the sector.
## BACKGROUND
AI agents are autonomous software systems powered by foundation models that can perceive their environment, break down complex tasks, and execute multi-step actions on behalf of users without continuous human intervention. As generative AI models advance rapidly, researchers and regulators increasingly worry about existential risks, misuse in cyber operations or weapons development, and the economic fallout of unconstrained autonomous agents.