Anthropic Urges Global Pause on Powerful AI

Anthropic, maker of the Claude AI models, has called for a worldwide slowdown in cutting-edge AI development, warning that the most powerful systems are showing signs of escaping human control. The company says a pause would only work if major AI players in both the US and China agreed simultaneously under verifiable rules, comparing the challenge to nuclear arms control. The proposal faces significant resistance in Washington and Silicon Valley, where officials argue any slowdown risks ceding ground to China. Anthropic also warned that AI is already accelerating its own development, creating a feedback loop that could lead to recursive self-improvement sooner than governments are prepared for.