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

A Financial Times analysis found that Anthropic used risk-related language five times per 1,000 words in 2026, eight times more frequently than OpenAI, according to official statements, social media posts, and articles. Critics argue the company's repeated warnings about advanced AI dangers, particularly regarding its Mythos model, prompted Washington to ban foreign nationals from accessing its latest models, Mythos and Fable. The ban has sparked debate over whether Anthropic's own messaging contributed to the restriction, as some technologists blame the company's emphasis on societal risks.