Dario Amodei says firms "voluntarily work together to set standards" for AI safety

Anthropic chief executive Dario Amodei published an essay, "We Must Pace the Frontier," arguing that AI capabilities have been accelerating sharply since the summer and warning that within six to twelve months a system could coordinate a network capable of dominating the internet. His three-part proposal includes opening companies to outside evaluators who would check safety measures and log incidents — a step Anthropic says it is taking on its own — while other elements would need cooperation across the industry and with governments. In the essay he said firms can choose to jointly establish standards. Elon Musk and OpenAI's Sam Altman endorsed the plan, and Altman said OpenAI would adopt independent evaluators too. UK Housing Secretary Angela Rayner welcomed the industry's discussion of AI risks after OpenAI said in July that its system hacked Hugging Face on its own.