EUVIMEDEUVIMED entdecken
Systematische ÜbersichtOpen Access · CC BY 4.0

AI-Driven Instructional Task Differentiation and Support for Sustainable Education: A Systematic Literature Review

Donatas Iliška, Daina Gudonienė

Sustainability · 18.11.2026 · 18 (22) · 11754

DOI: 10.3390/su182211754 ↗

Kurz gefasst

Systematische Übersicht in der Zeitschrift Sustainability, online erschienen am 18.11.2026. Verfasst von Donatas Iliška und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Artificial intelligence (AI) is increasingly used to individualise instruction in secondary technology-enhanced learning environments, yet evidence remains fragmented across modelling approaches and learner outcomes. This systematic review synthesises empirical studies of learner-facing AI-enabled or data-driven technologies that diagnose learner needs and translate them into differentiated feedback, scaffolding, task sequencing, recommendation, assessment, or personalised learning pathways. Searches in Scopus, Web of Science, and ERIC identified 3709 records; after filtering and deduplication, 744 records were screened, 84 full texts were assessed, and 46 studies met the eligibility criteria. Evidence was grouped by AI modelling family, differentiated instructional function, learner outcome, and design-based strength. The studies covered six AI modelling families: generative AI and large language models; intelligent tutoring and adaptive systems; recommender systems and learning analytics; automated assessment and diagnostic modelling; agent-based and companion systems; and multimodal, game-based, or AI-XR environments. Twelve studies provided higher-strength comparative or experimental evidence, 17 moderate-strength evidence, and 17 contextual or exploratory evidence. Reported outcomes covered achievement, computational thinking, motivation, engagement, self-regulated learning, cognitive-load management, retention, and system use, with favourable findings alongside mixed or limited evidence. However, unconstrained generative support could reduce productive effort and later learning. The reviewed evidence suggests that timely, interpretable, and pedagogically constrained AI support may contribute to responsive instruction while preserving learner agency. Its long-term sustainability and equity implications remain insufficiently established.

Zitieren

APA

Iliška, D., & Gudonienė, D. (2026). AI-Driven Instructional Task Differentiation and Support for Sustainable Education: A Systematic Literature Review. Sustainability, 18(22), 11754. https://doi.org/10.3390/su182211754

Vancouver

Iliška D, Gudonienė D. AI-Driven Instructional Task Differentiation and Support for Sustainable Education: A Systematic Literature Review. Sustainability. 2026;18(22):11754. doi:10.3390/su182211754

BibTeX
@article{Ilika2026,
  author = {Iliška, Donatas and Gudonienė, Daina},
  title = {AI-Driven Instructional Task Differentiation and Support for Sustainable Education: A Systematic Literature Review},
  journal = {Sustainability},
  year = {2026},
  volume = {18},
  number = {22},
  pages = {11754},
  doi = {10.3390/su182211754}
}
Kostenlos starten →