Anthropomorphism in Children's Interactions with LLM Chatbots

A systematic review of 35 empirical studies (2022-2025) identifies how children anthropomorphize LLM chatbots, finding that human-like personas, adaptive scaffolding, supportive companionship, and non-human design drive these interactions. The research reveals five key outcomes: children exhibit paradoxical social and moral responses, maintain dual consciousness about the chatbots' nature, form varying social bonds, explore social boundaries, and attribute human narratives to system failures. These findings highlight both benefits and risks that should inform the ethical design of child-focused LLM chatbots.