Governing Children’s and Adolescents’ Exposure on Social Media: A Taxonomy of Risks and Harms

Public debate on children’s safety in social media has focused mostly on age verification. While important, it regulates who can access platforms, not what children and adolescents are repeatedly exposed to. Yet, harms may arise from isolated age-inappropriate content as well as from cumulative exposure over time. Protecting minors therefore requires broadening the debate beyond access control to include exposure governance. This shift raises major research challenges, including the need for novel context-informed risk taxonomies and computational frameworks for modeling exposure, supported by advances in machine learning, natural language processing, temporal modeling, causal inference, privacy-preserving systems, and interdisciplinary collaboration with other fields. If computing helped shape children’s digital environments, it must now help govern what algorithms repeatedly place before young minds.