아동중심 AI 기술의 미래의제 도출 연구

Abstract

This research aims to examine the current practices and issues of AI use in childcare and identify the key topics concerning the application of AI technology from a child-centered perspective in the 0-12 age group (childhood education, care, and parenting). This study is structured around four main research scopes and employs a mixed-methods approach to comprehensively investigate the application and future direction of Child-Centered AI technology. To achieve these goals, the research employs a robust methodology that includes: literature review, on-site visits to kindergartens and daycare centers for AI application case studies, social big data analysis, various forms of interview investigations (FGI) with both demand and supply groups (industry personnel), a large-scale online survey of parents, and formal expert surveys and advisory meetings. As a result, it should be proposed and established that Child-Centered AI Policy and Guidelines, moving beyond defining children merely as passive subjects requiring protection and instead redefining them as active agents capable of realizing their rights within the digital environment. To this end, it is critical to immediately establish a system that explicitly includes children as a distinct protection target, placing their rights at the center of the policy framework. Specific legislative actions should include the legalization of age-based design principles (Design for Age), ensuring age-appropriate information accessibility and safety from the initial service design phase. Furthermore, a 'customized guidance system' must be built to provide clear, understandable information to both children and their guardians. It is essential to institutionalize the responsibility of developers and companies to design based on children's rights and actively incorporate platform companies' self-regulatory examples (e.g., voluntary adoption of child protection measures) into policy-making. The future agenda for child-centered AI technology is structured around guaranteeing and advancing five key dimensions: Development, Rights (Protection/Participation), Competency, Agency, and Fairness, all of which must be tailored to a child's age and developmental requirements. The core implementation tasks center on several critical areas. Firstly, achieving Child Rights and Agency necessitates guaranteeing child rights and participation when implementing AI in the parenting domain. Secondly, fostering Child Competency and Literacy requires implementing a process of Co-agency, where children's developing competencies (like creativity and literacy) engage in mutual collaboration with the autonomy and evolving sophistication of AI. Other crucial tasks involve ensuring AI technology is Child-Friendly and Inclusive by applying appropriate levels and paces of development, and improving Data Utilization and Accessibility by creating data-driven support for vulnerable children, developing strategies to prevent the AI access gap, and building systematic data infrastructure for the parenting sector. Furthermore, the agenda focuses on Establishing a robust AI System for Education and Care through supporting AI introduction and teacher readiness, alongside monitoring implementation. Finally, the agenda addresses governance by Securing AI Regulation and Human-Centricity through institutionalizing AI control and child-centric norms, and Responding to Socio-Economic Structures and Inequality by establishing systems to share and mitigate the risks of inequality generated by AI in the parenting field.