Artificial Intelligence-Based Interventions to Enhance Joint Attention Skills in Children with Autism Spectrum Disorder: A Systematic Review

Abstract

Background: Joint attention (JA) is a key component of social, cognitive, and language development, yet children with autism spectrum disorder (ASD) often experience difficulties with it. Despite its crucial role in social interaction and communication, therapeutic interventions have rarely targeted JA directly. With advances in artificial intelligence (AI), particularly in autism research, new methods have emerged to enhance JA and social skills in this population. However, studies evaluating their effectiveness and characteristics remain scattered, highlighting the need for a comprehensive systematic review to analyze the available evidence.
Method: This systematic review followed the PRISMA 2020 guidelines. Major databases (Scopus, IEEE Xplore, PubMed, ERIC, ProQuest) and additional sources were searched using specific and advanced search terms, covering the period 2015–2025, to identify studies on AI-based interventions for improving JA in children with ASD. The search yielded 556 records, of which only 11 met the inclusion criteria after screening.
Results: Most AI-based interventions demonstrated moderate to high effectiveness, particularly in controlled environments. The majority of interventions primarily relied on humanoid robots. Nonetheless, variations in intervention protocols, small sample sizes, and short follow-up periods limited the generalizability of findings.
Conclusion: AI-based interventions show promising potential for improving joint attention in children with ASD. However, high-quality studies are urgently needed to assess their long-term effectiveness, applicability in natural settings beyond experimental conditions, and ethical considerations during use.