Artificial Intelligence‐Enabled Feedback for Reflective Learning Across Health Professions Education: A Scoping Review

ABSTRACT Although artificial intelligence (AI) is increasingly used to support feedback in health professions education, little is known whether and how AI‐enabled feedback supports reflective learning and critical reflexivity. The literature encompasses generative large language models (LLMs) as well as non‐generative natural language processing (NLP), traditional machine learning (ML) and computer vision (CV). This scoping review examines the design and use of AI‐enabled feedback for reflective learning across health professions education. Following Arksey and O'Malley, Levac et al. and PRISMA‐ScR guidance, we searched MEDLINE, Embase and ERIC from 1 January 2015 to 21 November 2025. Eligible empirical studies examined AI‐enabled feedback or feedback‐related educational activities involving health professions learners or professionals in undergraduate, graduate, or continuing professional development contexts. Data were charted descriptively and classified using FACET, SAMR, Hattie and Timperley's feedback framework and Kirkpatrick outcome levels. Nineteen studies met the inclusion criteria. A total of 12 of 19 (63.2%) used LLMs, 4 (21.1%) used non‐generative NLP, 2 (10.5%) used traditional ML and 1 (5.3%) used CV. Six studies (31.6%) involved continuing professional development (CPD). Augmentation was the most common SAMR level (10/19; 52.6%). Of 15 studies that could be classified by Hattie and Timperley's framework, 10 (66.7%) provided process‐level feedback, 3 (20.0%) task‐level feedback and 2 (13.3%) self‐regulation feedback. No included study explicitly examined critical reflexivity. Reported outcomes were concentrated at Kirkpatrick Level 1 (reaction; 9/19) and Level 2 (learning; 9/19), with no demonstrated longitudinal evidence of sustained behaviour change, transfer to practice, or lifelong learning outcomes. Current evidence on AI‐enabled feedback in health professions education is early, fragmented and predominantly performance‐oriented. Critical reflexivity has not yet been adequately studied, and lifelong learning should be regarded as a future educational goal rather than a demonstrated outcome of the included evidence. Current evidence on AI‐enabled feedback is predominantly performance‐oriented, with limited examination of reflective learning and no explicit evaluation of critical reflexivity. Although AI‐enabled feedback may contribute to ongoing professional development, its effects on sustained learning, behaviour change and lifelong learning remain unestablished. Future work should use theory‐informed feedback and design, strengthen learner feedback literacy and human oversight, address bias and power, and evaluate meaningful educational and practice‐based outcomes over time.

Authors

Institutions

Publication Details

Journal
European Journal of Education
Published
2026-11-14
DOI
https://doi.org/10.1111/ejed.70938
Primary Topic
Artificial Intelligence in Healthcare and Education
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Artificial Intelligence‐Enabled Feedback for Reflective Learning Across Health Professions Education: A Scoping Review

Alice Sau Han Kam, S. Yu, Ayaan Islam, Siyona Kassel et al.
European Journal of Education
Artificial Intelligence in Healthcare and Education
article

Artificial Intelligence‐Enabled Feedback for Reflective Learning Across Health Professions Education: A Scoping Review

Alice Sau Han Kam, S. Yu, Ayaan Islam, Siyona Kassel, Ronny Thomas
article en

Abstract

ABSTRACT Although artificial intelligence (AI) is increasingly used to support feedback in health professions education, little is known whether and how AI‐enabled feedback supports reflective learning and critical reflexivity. The literature encompasses generative large language models (LLMs) as well as non‐generative natural language processing (NLP), traditional machine learning (ML) and computer vision (CV). This scoping review examines the design and use of AI‐enabled feedback for reflective learning across health professions education. Following Arksey and O'Malley, Levac et al. and PRISMA‐ScR guidance, we searched MEDLINE, Embase and ERIC from 1 January 2015 to 21 November 2025. Eligible empirical studies examined AI‐enabled feedback or feedback‐related educational activities involving health professions learners or professionals in undergraduate, graduate, or continuing professional development contexts. Data were charted descriptively and classified using FACET, SAMR, Hattie and Timperley's feedback framework and Kirkpatrick outcome levels. Nineteen studies met the inclusion criteria. A total of 12 of 19 (63.2%) used LLMs, 4 (21.1%) used non‐generative NLP, 2 (10.5%) used traditional ML and 1 (5.3%) used CV. Six studies (31.6%) involved continuing professional development (CPD). Augmentation was the most common SAMR level (10/19; 52.6%). Of 15 studies that could be classified by Hattie and Timperley's framework, 10 (66.7%) provided process‐level feedback, 3 (20.0%) task‐level feedback and 2 (13.3%) self‐regulation feedback. No included study explicitly examined critical reflexivity. Reported outcomes were concentrated at Kirkpatrick Level 1 (reaction; 9/19) and Level 2 (learning; 9/19), with no demonstrated longitudinal evidence of sustained behaviour change, transfer to practice, or lifelong learning outcomes. Current evidence on AI‐enabled feedback in health professions education is early, fragmented and predominantly performance‐oriented. Critical reflexivity has not yet been adequately studied, and lifelong learning should be regarded as a future educational goal rather than a demonstrated outcome of the included evidence. Current evidence on AI‐enabled feedback is predominantly performance‐oriented, with limited examination of reflective learning and no explicit evaluation of critical reflexivity. Although AI‐enabled feedback may contribute to ongoing professional development, its effects on sustained learning, behaviour change and lifelong learning remain unestablished. Future work should use theory‐informed feedback and design, strengthen learner feedback literacy and human oversight, address bias and power, and evaluate meaningful educational and practice‐based outcomes over time.

European Journal of EducationVol. 61(4)
University of Toronto (CA)
Artificial Intelligence in Healthcare and Education
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon