From the course: The OWASP Top 10 for Large Language Model (LLM) Applications: An Overview
Join today to access over 24,900 courses taught by industry experts.
Misinformation mitigations
From the course: The OWASP Top 10 for Large Language Model (LLM) Applications: An Overview
Misinformation mitigations
- [Instructor] In the last video, we talked about what the misinformation vulnerability is all about. Now, let's discuss how to reduce this particular risk for large language models. First, connect your large language model to trusted and verified sources using retrieval augmented generation, or RAG, which we covered earlier in the course. This means that the LLM does not just rely on what it learned during training. Instead, it also pulls real-time information from reliable databases or documents to provide accurate answers, helping reduce hallucinations and misinformation. For example, a healthcare chatbot connected to verified medical databases will give safer advice than one relying solely on the internet text. Second, build in human review and fact-checking in your process. LLMs can speed things up, but when the output affects real people, like in healthcare, legal, HR, or financial settings, a trained human should always review the results. Think of it like a doctor reviewing…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.