Privacy policy clauses for LlamaIndex
LlamaIndex is a data framework that indexes and retrieves document content to enhance large language model (LLM) applications. Websites use it to enable retrieval-augmented generation (RAG), allowing AI systems to access and reference specific documents when answering user queries.
Free scan · No signup · Results in 60 seconds
What data LlamaIndex collects
Your privacy policy must disclose each of the following data types when you use LlamaIndex.
When does LlamaIndex trigger privacy obligations?
LlamaIndex triggers privacy obligations the moment you start indexing document content and routing user queries through it. Three data flows activate immediately:
1. Document indexing: Your application uploads or references documents (PDFs, web pages, databases) into LlamaIndex's indices. This constitutes collection and processing of personal data if those documents contain identifiable information. Under GDPR Article 13/14, you must disclose this processing in your privacy notice.
2. Query processing and embeddings: User searches are converted into embeddings and stored. If users are identifiable (logged-in users, email queries), GDPR Articles 5–6 require lawful basis and data minimization.
