Determine children’s firsthand experiences with AI toys

Determine how children experience AI toys that incorporate large language models for open-ended spoken dialogue, memory across interactions, and personalized responses in practice from the child’s firsthand perspective during play and daily use.

Background

The paper notes that most existing work and public discourse about AI toys emphasize adult perspectives, policy concerns, or technical safety risks, with comparatively little empirical study of children’s own interactions. AI toys built on LLMs introduce open-ended dialogue, personalization, and persistence, potentially reshaping play and social expectations.

Given this gap, the authors foreground the need to directly investigate children’s firsthand experiences with these toys in real interactions, motivating their participatory design study with Curio’s conversational AI toys (Grok, Grem, and Gabbo).

References

Much of the emerging research has focused on adult perspectives, policy implications, or technical safety risks, leaving open questions about how children experience these systems in practice first hand.

— Toys that listen, talk, and play: Understanding Children's Sensemaking and Interactions with AI Toys  (2604.02629 - Dangol et al., 3 Apr 2026) in Section 1 (Introduction)

Two rival explanations remain. Callers had dialed a Santa hotline and may have disengaged when an assistant was not who they expected, and the personas differed in voice as well as identity. We cannot separate deference to authority from either.

— Persona and Persuasive Framing in AI Voice Agents: A $2\times2$ Field Experiment with Children  (2609.38782 - Tamme et al., 30 Sep 2026) in Section 5.1, Discussion

This raises a question about whether the system could accommodate individualized safety guardrails and interaction styles.

— Understanding Parents' Complex Views of AI for Children's Pretend Play  (2609.39906 - Oh et al., 30 Sep 2026) in Section 5.1, “Parental Protection for Child Safety”