A Generative AI Engineer is developing a chatbot to help users with insurance-related questions. The chatbot is conversational and powered by a large language model (LLM). However, to ensure the chatbot stays on topic and adheres to company policy, it should not answer any political questions. Instead, when faced with political inquiries, the chatbot should respond with the standard message: “Sorry, I cannot answer that. I am a chatbot that can only answer questions around insurance.” What type of framework should be implemented to achieve this?