This article discusses the emerging challenges posed by artificial intelligence to the protection of children and adolescent’s personal identities. The specific characteristics of personal identity of children and their vulnerability in the digital environment are examined, where AI systems can generate new types of harm that transcend traditional categories of private law, as well as some specific types of harm caused by the use of AI. The complex challenges for the attribution of civil liability are analyzed, considering the multiplicity of actors involved and the existence of evidentiary difficulties. The study concludes with proposals on the need to adapt the existing regulatory frameworks.