JUST IN: 🇺🇸 Anthropic proposes new AI safety law to US lawmakers, allowing the government to block dangerous AI models.