Assessing the Age of Unaccompanied Foreign Children and Adolescents with AI: New Innovations, Old Dilemmas

Abstract The debate on the best way ethically to assess age in unaccompanied migrant adolescents, particularly for access to child protection in the realm of migration management, is decades old but still relevant. As such, innovative ways to provide an accurate result with forensic methods includes the use of AI for the generation of up-to-date, extensive databases that can more objectively compare radiographic images. In this article, we reflect on old and new approaches in age assessment. We consider the large margin of error in using biological age as a proxy for chronological age and the motivation of using age as a filter for access to child protection. We conclude that forensic methods are a way to organise entry into (or exclusion from) protection systems, but cannot be considered – on their own – methods that maintain the best interests of the child.