Redefining personalized learning in the artificial intelligence era: an updated systematic review from 2019 to 2025

Synthesis: Khalifeh, Santiago, and Palau (2026) systematically review how AI is redefining personalized learning (PL) terminology from 2019–2025 across six databases. They find that PL remains a broad, conceptually ambiguous umbrella term — with adaptive learning, individualized instruction, and customized learning used interchangeably — and argue that AI-driven real-time feedback, adaptive sequencing, and intelligent tutoring have transformed PL in ways that demand a unified framework and definition. They also identify persistent research gaps: a disproportionate focus on older learners, limited longitudinal/inclusive/methodologically rigorous research, and insufficient attention to privacy, bias, Ethics, teacher preparation, and human–AI collaboration.

Key Findings

1. PL is conceptually ambiguous. Personalized learning functions as a broad, multi-layered umbrella term without a universally accepted definition; related terms are used interchangeably, creating clarity problems for research and practice.
2. AI is transforming PL. Real-time feedback, adaptive sequencing, intelligent tutoring, and dynamic learning pathways have shifted PL from static, one-size-fits-all instruction toward adaptive, data-driven, learner-centered approaches.
3. A unified term is needed. The review calls for a common framework to clarify definitions and enhance consistency across educational technology practice — extending Shemshack and Spector's (2020) earlier terminology review.
4. Research gaps persist. Disproportionate focus on older learners; a lack of longitudinal, inclusive, and methodologically rigorous research, especially in underrepresented populations and diverse contexts; frameworks lagging behind technological innovation; and under-addressed concerns around privacy, bias, ethics, teacher preparation, and human–AI collaboration.

What this means for practice

• Designers. Stop shipping "personalized" as a capability claim: the review finds personalized learning to be an umbrella term with adaptive learning, individualized instruction, and customized learning used interchangeably, so state which mechanism — adaptive sequencing, real-time feedback, or intelligent tutoring — a product actually implements.
• Designers. Separate adaptivity from personalization in product documentation: AI-driven adaptive sequencing is one design family in the literature, not a synonym for tailoring instruction to a learner.
• Researchers. Adopt a shared definition before comparing studies: the review's central recommendation is a unified framework and definition, because inconsistent terminology is what keeps findings from aggregating.
• Researchers. Fill the gaps the review documents rather than adding another short-term study of older learners: longitudinal, inclusive, and methodologically rigorous work in underrepresented populations is what the field lacks.
• Administrators. Fund teacher preparation and human–AI collaboration alongside the technology, since Privacy, bias, Ethics, and teacher readiness are named as insufficiently addressed in the literature.

Limitations

• As a systematic review, the findings are bounded by its source base — literature from 6 databases over 2019–2025 — and it analyzes terminology and concepts rather than pooling effect sizes.
• It cannot establish which personalized-learning approach improves learning: the emphasis is conceptual and terminological, and empirical validation of the proposed unified definition remains to be done.
• The review documents the field's own limits — little longitudinal, inclusive, or methodologically rigorous research and disproportionate attention to older learners — which caps the strength of any conclusion drawn from the body of work it synthesizes.

Citation

Khalifeh, F., Santiago, R., & Palau, R. (2026). Redefining personalized learning in the artificial intelligence era: an updated systematic review from 2019 to 2025. Smart Learning Environments, 13, 19. https://doi.org/10.1186/s40561-026-00440-6