<![CDATA[Coding and artificial intelligence (Coding-AI) are increasingly introduced in elementary schools to provide age-appropriate experiences in computational thinking, purposeful technology use, and responsible engagement with artificial intelligence. However, evidence on how Coding-AI is implemented as an additional subject in Indonesian private elementary schools remains limited. This qualitative case study examined the implementation of Coding-AI in one Grade-5 class at SD Muhammadiyah 1 Kesamben, Blitar, Indonesia, during the 2025/2026 academic year. Participants were one teacher and 15 students. Data were generated through classroom observations, semi-structured interviews, and documentation and analyzed using the interactive processes of data condensation, data display, and conclusion drawing/verification. The analysis used six sensitizing competence elements: computational thinking, digital literacy, algorithms and programming, data analysis, AI literacy and ethics, and AI use for learning. Case-specific coded evidence was documented for digital literacy among 15/15 students (100.0%), computational thinking among 14/15 (93.3%), AI literacy and ethics among 14/15 (93.3%), algorithms and programming among 13/15 (86.7%), AI use for learning among 13/15 (86.7%), and data analysis among 12/15 (80.0%). These frequencies summarize observed or reported behaviors within the case and do not represent standardized competency scores or changes over time. The teacher facilitated stepwise problem solving, purposeful and safe technology use, introductory programming, and ethical AI exploration, while students described systematic task completion, information seeking, and guided engagement with digital learning resources. The convergence of classroom, interview, and documentary evidence informed a contextual representation of Coding-AI implementation that emphasizes scaffolded participation rather than independent mastery. Because the findings derive from one teacher and 15 Grade-5 students in one private school, they should be interpreted as analytically, rather than statistically, transferable.]]>
