Design a Retrieval-Augmented Generation (RAG) system to help TikTok’s content moderation team detect and handle misinformation.

To design a RAG system for TikTok's content moderation, we need to address the detection and handling of misinformation. The system would integrate a retrieval component to access a knowledge base of verified facts, trusted sources, and historical misinformation patterns. When a piece of content is flagged, the retrieval system would fetch relevant, factual information. This retrieved context would then be fed into a generation model (the 'brain') which, acting as a reasoning engine, would compare the content against the retrieved facts to determine if it constitutes misinformation (e.g., false claims, twisted narratives, or incomplete information). The system's goal is to enhance platform reliability and user trust, ultimately fostering long-term engagement with accurate informational content.