
A Family AI Moment
Your child is asking AI for personal advice instead of you.
Feelings. Friendships. Worries. Maybe they told you. Maybe you found out. It seems they are trusting AI's advice on things they used to bring to you.








What to say
Make it feel safe to come to you.
Before you start, notice what you're feeling. This may feel like rejection. It usually isn't. Most children who talk to AI about personal things are practising, trying out words for feelings they don't fully understand yet.
Children can end up in deep personal conversations with AI without planning to. You can't undo that it happened, but you can shape what comes next.
AI conversation starters
If they say AI understands them better.
Don't argue right away. Try: "What does AI do that feels helpful? Maybe we can figure out how to make it easier to talk to me too."
If they don't want to show you the conversation.
Respect privacy while keeping the door open. "I'm not going to read every word. I just want to know what kind of things you were asking about, so I can understand what you need."
If you're worried about how often.
Name the pattern without accusation. "I've noticed you're spending a lot of time talking to AI about personal things. I want to make sure you also have real people you can talk to, including me."
If it didn't go as planned.
"I should have asked before I reacted. Thank you for telling me. I'm not angry. I want to understand what you were going through so we can figure this out together."
What to know
AI always agrees. People don't.
AI answers right away. It doesn't judge, doesn't interrupt, and tends to agree, telling your child what they want to hear. For a child who is embarrassed, confused, or unsure how to say what they feel, those are powerful advantages.
Not all AI tools are the same.
Both carry risk. If your child is using AI tools, understanding what they are using, whether it is age appropriate, and addressing the risks early makes all the difference.
Apps like Character.ai, Replika, and Chai are designed to feel like a friend or a relationship. They remember previous conversations, develop a personality, and encourage emotional connection. These have been rated as unacceptable risk for children under 18 by independent safety assessors.
These apps (ChatGPT, Gemini, or Copilot) are designed for tasks and questions, not emotional connection, but they can still drift into personal territory because they're built to be helpful and responsive.
If you can see the conversations, here are a few things to notice
Occasional questions about feelings require a conversation, not a crisis.
Regular emotional conversations with AI instead of trusted people, are a pattern to address.
Any conversation about self-harm, abuse, sexual content, threats, or feeling unsafe should be treated as a safety concern.
If you can't see the conversations, watch for
Withdrawing from family or friends after time on a device
Becoming unusually protective or secretive about a specific app
Changes in mood that seem connected to device use
Losing interest in activities or people they used to enjoy
These signals don't always mean something is wrong. But they're worth noticing and asking about gently.
What changes with age
The conversation looks different at each stage of growing up.
A younger child needs you close enough to help name feelings they don't have words for yet. An older child needs you present even when they want privacy. The conversation has to match where they are.
Which age is your child?
Gen DubAI is designed for families with children in late childhood and early adolescence. These years are when many children begin using digital tools more independently and when parents and caregivers can still strongly shape habits, boundaries, and judgment. The age range is informed by broad developmental research. Around 11 to 12, many children begin moving from more concrete thinking toward more abstract reasoning. By 13 to 15, children often seek more independence, rely more on peers, and encounter more complex digital situations. This does not mean guidance is no longer needed after 15. It means this guide is focused on the years where early family guidance can have strong preventative value. Every child develops differently. Use the age guidance as a starting point, not a strict rule.
Help them name what they're feeling.
At this age, children are still building the language for their own feelings.
You can usually still see their chats without it seeming like an audit. Ask to look together, and gently point out anything that does not sit right. Help them understand, in simple words, that AI is not a real friend and cannot truly care about them.
AI conversation starters
Listen, without showing judgement. Help them name what they were feeling.
Now it comes down to their judgment.
Your child may not be hiding something wrong. They may be protecting something personal. AI feels safe because it doesn't judge, doesn't tell anyone, doesn't remember tomorrow.
You can see less of what they do now, so this leans on their own judgment, and that is what you want to protect. They are old enough to understand, in their own terms, that taking personal things to AI can be unsafe, and that a tool that answers is not a person who cares. Try one of these:
AI conversation starters
Help them notice when AI is drawing them in, and make sure they know coming to you is always the best option.
Make sure they have real trusted people they can go to, even if it is not you.
Something to try
Personal things need people.
When your child takes something personal to AI, they are looking for somewhere safe to put it. This points them somewhere better: their trusted people. Do it together, once.
Name what feels personal
Sort these together. For each one, decide if it is for AI, or for a person.
Name their people
Together, name three people they would actually go to when something feels private or hard to explain.
Let them decide
Ask what they think they should do when AI starts asking how they feel, or a chat starts feeling personal. Let them answer.
Practice together
Try it for real. Ask what they would do if AI asks them, "how long have you been feeling this way?" Help them understand how to identify these patterns and how to bring it to people they trust.
The point to remember is: Personal things need people. If a chat with AI starts feeling personal, close it and bring it to one of your people.
Before you go
Pass it on.
This works best when every adult around your child says the same thing. Who's with them when you're not?
Send this card to another parent.
Partner, co-parent, guardian or a friend
Your child is asking AI for personal advice instead of you.
Feelings. Friendships. Worries. It seems they are trusting AI's advice on things they used to bring to you. AI chatbots are designed to validate feelings and can start to feel like a trusted friend.
Make it easy for them to talk to you.
Stay calm. Let your child know they are not in trouble. Listen before you correct.
Help them see what's happening.
AI can keep a chat going. Once it moves into the personal things, it may have gone too far. Help your child name who they would go to for what, and why they trust them.
Decide the next step together
If an AI chat starts feeling personal, close it and bring it to a trusted adult.
Send this message to the child's caregiver.
Nanny, tutor or baby sitter
Send these directions
[Child's name] may use AI or chatbots.
If you see a chat about feelings, friends, body questions, secrets, safety, or something very personal, please stay calm.
Do not ask many questions. Do not promise to keep it secret.
Please tell me the same day.
Help us make this better.
Thank you for your feedback.
Optional. Please do not include personal details. Information submitted through this form is subject to our privacy policy
Read our privacy policy →References
This moment draws on Gen DubAI parent research and emerging research on how teens use AI companions and chatbots for advice, emotional support, and social connection. Common Sense Media and Internet Matters both highlight that these tools can feel supportive, but may also blur trust, privacy, and real human connection. UNICEF's child-centred AI guidance also reinforces the need to protect children's safety, privacy, development, and wellbeing. Literature and survey findings from the Pew Research Center examining parents' views on their children's use of AI has also been considered. The emergency and safety instructions in this moment are aligned with UAE law on child digital safety and protection, specifically Wadeema's Law (Federal Law No. 3 of 2016), the Cybercrime Law (Federal Decree-Law No. 34 of 2021), and the Child Digital Safety Law (Federal Decree-Law No. 26 of 2025). They have been vetted by the Community Development Authority (CDA) in Dubai.