Anthropic Calls for AI Slowdown, Warns Humans Could Lose Control of Technology

Anthropic, a leading AI company behind the Claude chatbot, has called for a temporary halt to developing more powerful AI models, arguing that humanity risks losing control of the technology. The company proposed the pause if other major AI developers agreed, citing the need for governments and researchers to better understand and manage emerging risks. In an essay published June 4, 2026, Anthropic’s research head Marina Favaro and president Jack Clark warned that AI could soon reach a stage of recursive self-improvement, where systems autonomously design and enhance future AI generations with minimal human input. The company noted AI is already accelerating its own development—employees now produce eight times more code than in 2021–2025 due to AI-assisted tools, while systems improve at generating ideas and scientific planning. Anthropic CEO Dario Amodei has previously estimated a 25% chance of severe negative outcomes if AI is not developed responsibly. The company has restricted access to its most advanced model, Mythos, over concerns it could enable large-scale cyberattacks or other misuse. Despite the warnings, Anthropic acknowledged a global pause would be challenging to enforce, given private AI training operations and competitive pressures akin to an arms race. Compliance verification would be difficult, as AI development often occurs behind closed doors. Critics argue Anthropic may be overstating risks to push for regulations favoring its position. The debate follows recent market volatility, including Broadcom’s stock decline after disappointing AI-related sales forecasts, raising questions about whether AI hype has outpaced reality.