Children’s ability to distinguish AIC responsiveness from relational subjecthood

Determine whether children and adolescents can distinguish an AI companion’s genuine responsiveness from the presence of a subject on the other side of the interaction, and establish at what ages they can hold this distinction.

Background

The paper argues that transparency about an AI companion’s configurative nature presupposes that users can distinguish between an AI system’s genuine responsiveness and the existence of an experiencing or relational subject behind that responsiveness. This distinction may be especially difficult for children and adolescents because their relational schemas and skills are still developing, and they may form mental representations of AI-companion relationships before acquiring the developmental capacities needed to understand and manage them.

The authors explicitly identify as unresolved whether children can maintain this distinction and at what developmental ages they can do so. Resolving this question is presented as relevant to the design of transparency practices, safety guardrails, and ethical protections for minors using AI companions.

References

Transparency about the system's configurative nature assumes a user who can hold the distinction between genuine responsiveness and a subject on the other side of it, and whether children can hold that distinction, and at what ages, is an open question.

— The AI Side of the Human-AI Companion Relationship: Contributions to Users' Mental Models  (2610.12891 - Fabes et al., 9 Oct 2026) in Section “Mental Models, AICs, and Design Tensions”