From Algorithmic Control to Learner-Driven Personalisation: Rethinking AI in Education

AI-driven personalised learning now shapes the pace, practice, and feedback of millions of learners. Most systems, however, personalise learning for the learner through opaque algorithmic decisions instead of equipping learners to personalise their own study. This review examines that asymmetry. Twenty-five studies published between 2017 and 2026, retrieved from Google Scholar, ERIC, and Scopus, are synthesised thematically around learner agency, self-regulated learning, and system design. The synthesis identifies an illusion of autonomy: surface-level choices sit alongside algorithmic authority over goals, sequencing, and mastery criteria, which narrows the room learners have to practise goal setting, strategic monitoring, and reflective evaluation. Duolingo and two platforms used in Vietnam serve as illustrative cases showing how gamified metrics and default pathways can redirect attention from meaningful goals towards externally driven engagement. The paper answers with a learner-driven personalisation framework resting on three principles: shared control, design that activates self-regulated learning, and algorithmic transparency coupled with AI literacy. A three-item rubric operationalises the framework so that schools can detect illusion-of-autonomy risks before adoption. The analysis extends to equity, since learners with limited devices, connectivity, or AI literacy gain the least from transparency measures that presuppose digital capital. Implications concern system designers, educational leaders, and teachers in Vietnam, who need evaluation criteria reaching beyond short-term performance. The contribution is a concise screening tool that links learner control, the three phases of self-regulated learning, and algorithmic transparency, complementing broader agency rubrics with a focused test for illusion-of-autonomy risks.

Authors

Publication Details

Journal
DS Journal of Multidisciplinary
Published
2026-11-19
DOI
https://doi.org/10.59232/dsm-v3i2p103
Primary Topic
Artificial Intelligence in Education
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

From Algorithmic Control to Learner-Driven Personalisation: Rethinking AI in Education

Vuong T. Pham, Minh Phan, Duong Thanh Linh
DS Journal of Multidisciplinary
Artificial Intelligence in Education
article

From Algorithmic Control to Learner-Driven Personalisation: Rethinking AI in Education

Vuong T. Pham, Minh Phan, Duong Thanh Linh
article en

Abstract

AI-driven personalised learning now shapes the pace, practice, and feedback of millions of learners. Most systems, however, personalise learning for the learner through opaque algorithmic decisions instead of equipping learners to personalise their own study. This review examines that asymmetry. Twenty-five studies published between 2017 and 2026, retrieved from Google Scholar, ERIC, and Scopus, are synthesised thematically around learner agency, self-regulated learning, and system design. The synthesis identifies an illusion of autonomy: surface-level choices sit alongside algorithmic authority over goals, sequencing, and mastery criteria, which narrows the room learners have to practise goal setting, strategic monitoring, and reflective evaluation. Duolingo and two platforms used in Vietnam serve as illustrative cases showing how gamified metrics and default pathways can redirect attention from meaningful goals towards externally driven engagement. The paper answers with a learner-driven personalisation framework resting on three principles: shared control, design that activates self-regulated learning, and algorithmic transparency coupled with AI literacy. A three-item rubric operationalises the framework so that schools can detect illusion-of-autonomy risks before adoption. The analysis extends to equity, since learners with limited devices, connectivity, or AI literacy gain the least from transparency measures that presuppose digital capital. Implications concern system designers, educational leaders, and teachers in Vietnam, who need evaluation criteria reaching beyond short-term performance. The contribution is a concise screening tool that links learner control, the three phases of self-regulated learning, and algorithmic transparency, complementing broader agency rubrics with a focused test for illusion-of-autonomy risks.

DS Journal of MultidisciplinaryVol. 3(2)
Artificial Intelligence in 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