From automation to augmentation: a bibliometric study of AI feedback in self-regulated learning

Abstract

Purpose: to analyze the development of research on AI-driven feedback in supporting self-regulated learning (SRL).

Method: This study applies a bibliometric approach combined with thematic analysis to examine research on AI-driven feedback in supporting self-regulated learning (SRL). Data were collected from the Scopus database for publications from 2010–2025, resulting in 328 selected documents. Bibliometric analysis using Scopus and VOSviewer was conducted to identify publication trends, collaboration networks, and keyword evolution, while thematic analysis explored the shift of AI’s role from automation toward augmentation in supporting learners’ self-regulation.

Findings: a rapid increase in research on AI-driven feedback for self-regulated learning (SRL), especially after 2021. The findings also indicate a shift from AI as an automation tool toward a learning partner that supports cognitive and metacognitive processes. However, further research is needed to develop more adaptive and personalized AI feedback systems for digital learning.

Implications: AI-driven feedback can enhance self-regulated learning by supporting cognitive and metacognitive processes and independent, reflective learning through adaptive, collaborative systems.

Originality: mapping the evolution of AI feedback in self-regulated learning, highlighting the shift from automation to augmentation, and integrating bibliometric and thematic analysis to identify trends, key contributors, and future research directions.