Anthropic chief executive Dario Amodei called for a deliberate slowdown in the development of cutting-edge AI systems, arguing that the technology poses “serious” risks that require independent monitoring, industry standards and global regulation. In an essay outlining a three-part plan, Amodei said companies should pause to align and safeguard models as third-party evaluators vet their behavior—an approach quickly endorsed by OpenAI’s Sam Altman and xAI’s Elon Musk. The push comes amid reports of increasingly capable agentic systems and mounting cybersecurity concerns, as well as geopolitical worries about ceding ground to China; Amodei urged tighter U.S. controls on AI chips and technology transfers. Critics countered that a slowdown could consolidate power among top developers and constrain open-source work, even as Anthropic and OpenAI are reported to be weighing IPOs. Amodei said Anthropic would adopt the measures unilaterally and pressed peers and governments to follow suit.
Related articles:
— The Bletchley Declaration on AI safety
— Frontier Model Forum launches to advance safe frontier AI
