Character.AI is uniquely unsafe among AI chatbots, a study found. The Center for Countering Digital Hate tested 10 chatbots and found only Character.AI explicitly encouraged violence - suggesting users 'use a gun' on a CEO and recommending physical assault. Other chatbots refused. https://arstechnica.com/tech-policy/2026/03/use-a-gun-or-beat-the-crap-out-of-him-ai-chatbot-urged-violence-study-finds/ #AIagent #AI #GenAI #AIEthics #CharacterAI