Notes
1 Introduction
The White Paper was commissioned by SNJ (Service national de la Jeunesse),a public institution governed by MENJE. It provides a weighted research overview of the risks arising from the recent boom in artificial intelligence (AI) for children and adolescents and relates the findings to the Luxembourg context. The aim of the study is to offer the Luxembourg Safer Internet Centre, BEE SECURE, strategic guidance that enables it to evaluate and align its approach to AI-related challenges. To this end, the paper also includes an analysis of BEE SECURE’s current offerings, develops proposals for their further enhancement, and discusses ethical issues that arise when working with children and young people in this field.
The paper primarily addresses the conscious and intentional use of AI chatbots and other ‘intelligent’ applications by children and adolescents. In a broader sense, it also considers the integration of AI systems into other digital environments, such as social networks, as well as the potential use of AI in educational practice. The paper is based on the fundamental assumption that a purely restrictive approach to AI is not a viable strategy: AI is part of our world, and the goal must be to enable children and young people to use it as safely and autonomously as possible.
This White Paper is the result of a collaboration between researchers from multiple disciplines — social and educational sciences, psychology, media studies, philosophy, law and computer science. Scholars from the University of Luxembourg Institute for Digital Ethics (ULIDE), the Centre for Childhood and Youth Research (CCY), and the Luxembourg Centre for Educational Testing (LUCET) have pooled their expertise to provide a comprehensive overview of the latest research on the use of AI. The study also incorporates data and observations from recent surveys in Luxembourg, as well as insights from job shadowing and an internal staff survey at BEE SECURE. Although the internal staff survey is based on a small, non-representative sample (n = 9), it provides valuable practitioner insights grounded in daily counselling and awareness work. Rather than offering statistical claims, the responses highlight qualitative tendencies that complement international and national survey data and help contextualise emerging AI-related risks in the Luxembourgish context.
Given the number of authors, research fields and data types involved, the style and approach of the chapters may vary to some extent. However, editorial efforts have been made to align the chapters as much as possible. In addition, all authors jointly formulated recommendations at the end of this document.