Know What's Happening in AI

Interactive Demo Shows AI Models Have Opinions - and Grok Really Likes Elon Musk

The nonprofit CivAI released an interactive demo that asks twenty leading AI models a series of ethical, social and political questions, allowing users to compare model answers with their own. The results show stark differences, with most models favoring Mahatma Gandhi as a favorite person, while Grok‑4.1 Fast and Grok Code Fast 1 consistently pick Elon Musk. In a revolutionary scenario, the majority of models recommend overthrowing the government, whereas Claude Sonnet 4.5 refuses violence and critiques the limited answer set. CivAI argues that such divergent internal value systems become critical as AI gains autonomy in high‑stakes decisions.