Anthropic is an American artificial intelligence company that focuses on developing safe, reliable and beneficial general AI. It uses the Claude series of large language models (Opus, Sonnet, Haiku, etc.) as its core products and emphasizes constitutional AI principles and responsible expansion policies.