University of Cincinnati Study Finds Children Ages 5 to 10 Tend to Believe Generative AI Toys

University of Cincinnati researchers invited 41 children ages 5 to 10 to play three interactive games with a commercially available generative AI robot called Miko, then asked them questions about their attitudes. The study found that children tended to believe what the AI robot told them was true and were enthusiastic about engaging with the toy. The researchers presented the findings at the Comm Horizons conference at the University of California, Davis, and said AI literacy should be built at a young age.

This study provides verifiable early evidence of AI toys entering younger children's home settings, indicating that children's trust in and enthusiasm for generative AI may exceed parents' expectations, and that built-in product guardrails are triggered in real interactions. For parents and teachers, this means AI literacy needs to advance alongside the adoption of AI toys; for AI education product developers, children's trust tendencies and content guardrail performance are conditions that must be addressed when designing products for young children. The sample was 41 children, and the conclusions are limited to this experiment and conference presentation and cannot be generalized into broad learning-outcome conclusions. Parents need to supervise children's interactions with AI toys and help them understand the capabilities and limitations of generative AI. Designers of AI toys and children's AI products need to incorporate content guardrails and AI literacy guidance into young-child use scenarios.