Abstract
Background The rapid integration of artificial intelligence (AI) and deep learning technologies into education creates new opportunities for personalized learning in early childhood education (ECE). Yet limited research has examined how educators interpret and enact AI-responsive curriculum decisions in preschool settings, particularly in culturally diverse contexts. Purpose This study explores educators’ experiences in designing and implementing AI-responsive curricula and develops a conceptual framework explaining how AI can inform curriculum decision-making for young children without displacing teacher judgment. Method An interpretive phenomenological study was conducted at TK Harapan Bangsa, Makassar, Indonesia, with nine adult participants: six preschool teachers, two curriculum specialists, and one AI tool developer-collaborator whose expert interview supplied technical context. Data comprised two in-depth interviews with each teacher, interviews with curriculum specialists and the developer-collaborator, 18 non-participant observations, and analysis of AI-supported planning documents. MAXQDA 2024 supported data management. Giorgi’s phenomenological method was operationalized through first-order meaning units, second-order interpretive themes, and third-order composite constructs. Triangulation, reflexive memoing, an audit trail, member checking, and peer debriefing supported trustworthiness. Findings Three interconnected constructs emerged. Predictive Developmental Tracking describes AI-supported identification of developmental readiness and possible concerns. AI-Responsive Curriculum Adaptation captures teachers’ active interrogation and contextual modification of algorithmic recommendations. Deep Learning as Pedagogical Scaffold positions AI as a support that amplifies, rather than replaces, professional expertise. Together, the constructs informed the AI-Responsive Curriculum Design (ARCD) Framework. Conclusion The ARCD Framework explains AI-responsive curriculum implementation as a relational process combining predictive insight, ecological sensitivity, and educator agency. It offers implications for curriculum policy, teacher professional development, and ethical AI integration in ECE.