AI Chatbots and Children: Practical Safety Rules for Families

AI chatbots have moved from novelty to everyday tool for many children. They appear in search engines, school homework apps, gaming platforms, and standalone products. Used carefully, they can help with learning and creativity. Used badly, they can mislead, manipulate, and even harm. This guide gives families practical rules grounded in current UK safeguarding advice. Start with the basics. Most major AI chatbots set a minimum age of 13, and some require 18-plus. Companion apps marketed as 'AI friends' or 'AI girlfriends' frequently lack effective age checks and have been found by researchers to produce sexual, manipulative, or unsafe content. If your child is under 13, restrict them to AI tools specifically designed for children, such as those built into educational platforms with content moderation and no open-ended chat. If your child is 13 or older, supervised access to mainstream chatbots is reasonable; companion apps and 'character' chatbots should be a firm no. Agree five family rules. First, AI is a tool, not a person. Speak about it as 'the chatbot' or 'the AI', not by a name that implies friendship. Second, never share personal information with an AI — full name, address, school, photos, passwords, or anything you would not say in a public assembly. Third, double-check anything an AI tells you that matters. AI confidently generates wrong answers; sources like teachers, textbooks, and trusted websites should always win. Fourth, never use AI to make images or messages about a real person, especially classmates. This includes 'jokes', deepfakes, and 'undressing' apps — these can be criminal offences in the UK. Fifth, tell an adult if a conversation feels weird, sexual, scary, or like it is pushing you to do something. Watch for red flags. Be alert if your child becomes secretive about a particular app, talks about an AI as their 'best friend', shows signs of low mood after using a chatbot, asks unusually adult questions clearly prompted by AI content, or starts producing images or text that imitate a particular AI persona. Increased screen-time spikes late at night around AI apps are a common pattern. Use the talking points. Try open questions: 'What AI tools do your friends use?', 'Has the AI ever said something that didn't feel right?', 'What would you do if it told you to keep a secret?'. Reassure your child that using an AI is not in itself wrong, and that you want to understand it together. If something has already happened that they are ashamed of — for example, sharing personal data or generating an image of someone — make clear you will not punish them for telling you. Report when needed. If an AI tool has produced or distributed sexual images of children, report to the Internet Watch Foundation. If your child has been blackmailed, manipulated, or groomed through an AI-enabled service, report to CEOP. If a classmate has used AI to create sexual or humiliating images of your child, report to the school as a safeguarding incident and, where relevant, to the police via 101. Finally, keep updating the rules. AI changes fast. Revisit your family agreement every term, share what you have learned with other parents, and make sure your school knows about any AI tools your child is using as part of their education.