Dr. Arosha Mediwake is an accomplished educator, researcher, and multimedia specialist with over 20 years of experience in information and communication technology (ICT) education. She holds a Ph.D. in Information Technology and has dedicated her career to advancing digital learning, multimedia education, artificial intelligence in education, and inclusive teaching methodologies. Dr. Mediwake currently serves as a Visiting Lecturer at several leading higher education institutions in Sri Lanka, including CINEC Campus, Kingston University (via ESOFT), the University of Moratuwa, the National Institute of Education, and AIMS Campus. Her research focuses on AI-enhanced learning environments, multimedia-based education, hypermedia, and educational technologies for slow learners. A two-time National Gold Medalist for Teaching Innovation, she has authored educational books, developed digital learning platforms, and led nationwide teacher training initiatives. Through her ICTBUS online platform, YouTube channel, television programs, and community outreach activities, she continues to promote innovative, technology-driven education and empower students and educators across Sri Lanka.

Leaving No Child Behind: Empowering Invisible Slow Learners through AI-Based Activity-Centered Learning in Sri Lankan Primary Education Abstract Education should empower every child; however, highly competitive education systems often benefit only those who can successfully adapt to standardized learning environments. In Sri Lanka, academic competition begins during the first years of schooling, driven largely by preparation for the Grade Five Scholarship Examination. From Grade 1 onwards, many children experience continuous pressure from schools, parents, and private tuition to perform in an examination-oriented culture. While this system creates opportunities for academically high-achieving students, it unintentionally marginalizes a significant group of children who learn at a slower pace. These "invisible slow learners" are neither identified as students with special educational needs nor able to compete effectively within conventional classrooms. Consequently, they frequently experience repeated academic failure, declining self-esteem, reduced classroom participation, school avoidance, irregular attendance, and a growing belief that education is not meant for them. At the same time, many of these children spend increasing amounts of time using mobile devices for entertainment rather than learning, widening the gap between their digital engagement and classroom experiences. This study proposes an Artificial Intelligence (AI)-based Activity-Centered Learning Framework designed specifically to empower slow learners in Grades 1–5 in the Piliyandala Education Zone, Sri Lanka. Rather than replacing teachers, the framework positions AI as an intelligent learning companion capable of identifying individual learning patterns, adapting lesson difficulty, delivering personalized multimedia instruction, providing immediate feedback, and encouraging continuous progress through gamified educational activities. The proposed model transforms passive learning into an enjoyable, student-centered experience that values progress over competition. The research adopts a mixed-method design involving students, teachers, and parents. Data will be collected through questionnaires, interviews, classroom observations, attendance records, and AI-supported learning analytics before and after implementation. The study evaluates changes in learning motivation, engagement, confidence, classroom participation, foundational literacy and numeracy, and school attendance among underperforming learners. This research argues that the greatest promise of Artificial Intelligence in education lies not in producing higher examination scores but in restoring learning opportunities for children who have become invisible within competitive education systems. By providing personalized, equitable, and emotionally supportive learning experiences, AI has the potential to transform educational disadvantage into meaningful participation and lifelong learning. The findings are expected to offer an evidence-based framework for integrating AI into inclusive primary education and contribute to international discussions on educational equity, child well-being, and Sustainable Development Goal 4—ensuring inclusive, equitable, and quality education for all. Keywords: Artificial Intelligence in Education, Inclusive Education, Slow Learners, Educational Equity, Primary Education, Activity-Centered Learning, Adaptive Learning, Student Motivation, School Attendance, Grade Five Scholarship Examination, Sri Lanka.