Digital and AI-Supported Game-Based Interventions for Executive Function Development in Early Childhood: A Systematic Review

Abstract

Executive function (EF)—working memory, inhibitory control, and cognitive flexibility—underpins school readiness, academic achievement, and socio-emotional competence in early childhood (ages 3–8). Digital and AI-supported game-based interventions are a promising but underexamined modality for EF training. This review aimed to synthesize the reported effects of digital and AI-supported game-based interventions on executive function in children aged 3–8 years and the technological features linked to reported gains. Following PRISMA 2020 guidelines, this systematic review synthesized 19 studies from six databases (PubMed, PsycINFO, ERIC, Scopus, Web of Science, IEEE Xplore) covering January 2012–March 2026, involving 1890 children aged 4–8 years across 11 countries. Narrative synthesis indicated predominantly positive findings across all three EF components; this was observed most consistently for working memory. Adaptive difficulty calibration, immediate corrective feedback, and motivational scaffolding were the features most often linked to gains; positive findings were somewhat more frequent in school-based settings and in longer interventions, though these descriptive patterns rest on small, unbalanced subgroups and cannot establish setting or duration as moderators. Critically, the corpus was overwhelmingly conventional and rule-based: only three studies (15.8%) contained machine-learning components, with none driving pedagogical personalization (confined to gesture recognition or post hoc analysis); none employed conversational agents or affect-sensing, and no eligible study sampled 3-year-olds. Findings should therefore be read as evidence about adaptive digital game-based interventions rather than AI-driven ones. This near-absence of machine-learning personalization for typically developing children aged 3–8 years is itself a principal finding, alongside gaps in long-term retention, cross-cultural generalizability, and ethical AI data governance.

The authors' abstract, as published at the source. Education Sciences, 2026 · DOI ↗