LLMs for Explainable Artificial Intelligence: Automating Natural Language Explanations of Predictive Analytics Models

Abstract

This chapter explores the integration of LLMs into the Explainable AI (XAI) pipeline within the field of Learning Analytics (LA). The focus of the chapter is on how to automate the transformation of XAI outputs into accessible, natural language explanations using LLMs. Namely, we demonstrate how LLMs can contextualize feature importance, partial dependence profiles, and local explanations, making predictive model outputs more interpretable and actionable for non-technical stakeholders. We illustrate this process making use of an open source language model through the LM studio software. However, the model used shares the API definition with some widely used commercial models, such as those implemented by OpenAI, making the portability almost immediate. This work contributes to advancing the intersection of generative AI, XAI, and learning analytics to promote transparency, inclusivity, and fairness. © 2026 The Editor(s) (if applicable) and The Author(s).