Definition
Retrieval-augmented generation is a technique where an AI system looks up relevant documents at answer time and uses them to write its response, instead of relying only on what it learned in training. Data licensed for retrieval is read, not learned.
Why it matters when licensing company data
The distinction matters commercially: a retrieval license can be switched off (remove the documents and the model stops using them), while a training license leaves a lasting effect on the model. Retrieval deals are often priced per use or per period; training deals more often as a fee.
Source
Lewis et al., Retrieval-Augmented Generation (arXiv:2005.11401). Definition reviewed November 7, 2026 by Origin Data Partners.