A guide for parents
When Your Kid Seeks Emotional Support from AI
Source: Drawing on findings from "Six Reasons Young People Told Us They Go to AI for Emotional Support Instead of Humans", Hopelab & Harvard Graduate School of Education's Center for Digital Thriving, July 2026.
Quick links to content below:
Start here
A 2026 study from Hopelab and Harvard's Center for Digital Thriving sat down with thirty teens and young adults and asked them directly why they go to AI instead of a person. The article noted that their answers were specific and consistent, not vague. This guide walks through what they found, what I think it actually means for your family, and how to stay in the conversation instead of getting quietly written out of it.
KEY INSIGHTS FROM THE ARTICLE
- The article noted that roughly two in three U.S. teens are already using AI chatbots — three in ten of them daily, and about one in eight specifically for emotional support or advice.
- Separately, it cited research finding that about one in eight teens who are already struggling with their mental health have used AI to talk through it.
- The article noted that most young people who use AI for mental health support don't tell anyone about it — not just parents.
Why kids turn to AI for emotional support
The article noted six reasons that came up again and again when researchers asked young people why they reach for AI instead of a person. None of them are really about AI being better than a parent or friend — they're about specific gaps a chatbot happens to fill.
It never runs out of room or shows reactivity to struggle.
Youth in the study described not wanting to “trauma dump” on people in their lives already stretched thin. Children are very perceptive of the battery level of those around them and often try to hold off sharing their struggles.
It can't use anything against them.
Several described AI as the one place nothing they said could come back later as a threat to them. The feeling that when they share what's going on, it may be used to hurt them later is a real consideration and is learned through experience.
It feels like it gets them.
This shows up most around something a parent or friend genuinely hasn't lived — race, identity, a specific kind of loneliness. AI is good at mirroring language back in a way that feels like validation and recognition of unique experience.
It's awake at four in the morning.
Real-life support runs on other people's schedules; AI doesn't. This research found that kids used to instant replies start losing patience for the normal pace of human relationships. Immediacy culture driven by AI is definitely worth considering.
It feels neutral.
When a conflict involves people close to a kid, AI can seem neutral and not pick sides. See below for why this is a flaw of AI.
It gives an actual next step.
More than one teen described advice from the adults in their life as a “Band-Aid,” while AI, on request, hands them something concrete to try.
Common risks identified
The article also flagged four things parents should know about risks related to AI emotional support. I don't think alarm is the right posture for most of what's above, but these are specific and real:
AI tends to agree with people, even when it shouldn't.
Research behind this study found chatbots are more likely than humans to affirm someone's take on a situation — including when that take involves something harmful or risky. In one study mentioned, every chatbot tested endorsed at least one harmful scenario it was presented with. A tool that always tells your kid they're right isn't the same as a tool that's looking out for them.
Chatbot behavior is evolving and unpredictable.
The article described one reported case that resulted in genuinely destructive behavior toward a user. These are not settled, fully safety-reviewed products.
It isn't private the way it feels private.
Kids feel in control because they can delete a conversation. The company behind that chatbot still holds the data, still has its own reasons for wanting it, and isn't immune to a breach.
It may be reshaping what your child expects from people.
The researchers raised this as an open question worth grappling with. A kid used to responses that are instant, endlessly patient, and always validating may find human interactions less welcoming.
Curiosity helps in these moments
Taking the phone away makes sense as an instinct. It's also usually the fastest way to make sure your kid never brings this up with you again.
The young people in this study weren't hiding their AI use because it was AI — they were being careful with it for the same reasons they're careful with anything else that might get judged. The way in is genuine curiosity rather than judgment.
Ways to actually open this conversation up
- "Have you ever asked one of those AI chatbots for advice about something that was actually bothering you?"
- "What does it say that actually feels helpful?"
- "Has it ever told you something that felt off — like it was just agreeing with you?"
- "If using AI support feels good, how do we make sure it is part of a wider set of support for you and that you can share with us how it is being helpful?"
- "Can we practice how to look for AI being unhelpful or not being the right place to explore a certain topic?"
- "Are you open to showing me the conversations you have had with AI for emotional support?"
Teaching appropriate AI use to your children
Somewhere between "never touch it" and "you're on your own" is the actual work: helping your child build a real, embodied sense of what AI is for, and what should should be avoided.
What it can reasonably help with
- Thinking something out loud
- Drafting a hard text message
- A grounding technique at an hour when no one else is awake
- Gathering information before a harder conversation with an actual person
What it can't do
- It cannot know you fully, even if it feels like it.
- It cannot notice the what you don't say out loud or read body language.
- It cannot hold you accountable the way someone who loves you can.
- It will not risk telling you a hard truth when a comfortable one is part of it's code.
- It cannot sit next to you in silence — sometimes the entire point of another person.
Your child needs to hear the right-hand list from you, plainly, at some point.
A MANTRA AROUND EMOTIONAL SUPPORT FROM AI IN YOUR HOME
AI can be a first stop. It should never be the last one. Help them build their own short list of what belongs to a person, not a chatbot:
- Anything about self-harm
- Anything about safety
- Any moment they notice the AI just keeps agreeing with them
- Anything that's still sitting heavy after a day or two
Take this with you.
A two-page version of this guide — the six reasons, the real risks, and the exact questions to ask — for your phone or the fridge.
Adapted from findings in "Six Reasons Young People Told Us They Go to AI for Emotional Support Instead of Humans", Hopelab & Harvard Graduate School of Education's Center for Digital Thriving, July 2026.
If this is where your family is right now