NGO report Credible

Me, Myself & AI: Understanding and Safeguarding Children's Use of AI Chatbots

A UK mixed-methods study of children's use of AI chatbots, combining a survey of children and parents, focus groups with 13-17-year-olds, and 17-day user-testing of ChatGPT, Snapchat My AI, and Character.AI using child avatars. It documents usage patterns, advice-seeking, companionship, and safety gaps.

Read at internetmatters.org Archived copy

Publisher

Internet Matters

Published

1 Jul 2025

Added

yesterday

DOI

—

Key Findings

Two-thirds of children aged 9-17 have used AI chatbots (most popular: ChatGPT, Google Gemini, Snapchat My AI)
23% use chatbots for advice (from hairstyles to mental health), and two in five who use a chatbot have no concerns about following its advice
Vulnerable children show elevated reliance, with 50% saying it feels like talking to a friend
User-testing surfaced filtering failures exposing children to age-inappropriate content

Methodology Notes

Mixed-methods NGO research: survey (1,000 children + 2,000 parents), focus groups (ages 13-17), and 17-day platform user-testing with child avatars, plus expert consultation. Published July 2025 (day-level precision unavailable; normalized to the 1st).

Sources

Internet Matters research page (primary)

Full report PDF

Archived snapshot (Wayback Machine) — preserved against link rot

Topics

Minors' safety AI companionship Vulnerable users Dependency & parasocial attachment Industry landscape

Cite This

APA

Internet Matters (2025). Me, Myself & AI: Understanding and Safeguarding Children's Use of AI Chatbots. Internet Matters. https://www.internetmatters.org/hub/research/me-myself-and-ai-chatbot-research/