Anthropic's Frequent Warnings on AI Risks May Have Led to Export Ban

According to an FT analysis, Anthropic used risk-related words five times per 1,000 in its communications this year, compared to just 0.6 per 1,000 for OpenAI. The disparity has become politically charged after the US barred foreign nationals from using Anthropic's latest models, Mythos and Fable. Critics argue that the company's repeated warnings about AI dangers, especially concerning Mythos, contributed to the export restriction.