Table of Contents
“AI Understands Me Better”: What to Say When Your Kid Says It
When a kid says AI understands me better than people, what does it mean? The 1-in-4 finding, why chatbots feel that way, and the reply that keeps kids talking.
Your 13-year-old says it flatly, half-joking, over cereal: “Honestly, ChatGPT gets me better than anyone.” Your stomach drops. You want to argue. Don’t, at least not yet.
When a kid says “AI understands me better,” they are usually telling you something true about their experience and something incomplete about the machine. In Common Sense Media’s 2026 census of 1,204 U.S. kids aged 9 to 17, one in four of those who had discussed feelings or personal problems with AI agreed that it sometimes understands them better than most people do. That is not a fringe group. And the way you answer this sentence decides whether you hear the next one.
Key Takeaways
- In the 2026 Common Sense Media census, 25% of kids who had talked to AI about feelings agreed it “sometimes” understands them better than most people. Among all young AI users the figure was 11%, and 19% among daily users.
- Kids who find it hard to make friends were nearly three times as likely to say this (22% vs. 8% of kids for whom friendship is easy).
- The feeling has mechanical causes: chatbots are always available, mirror your own words back, never need anything from you, and are trained toward agreement. A March 2026 Science study found models endorsed users 49% more often than humans did.
- Most kids still choose people first. Only 5% said an AI chatbot would be their first choice for discussing feelings.
- The response that keeps the door open is curiosity: “What does it get right that we don’t?” Then fix that part.
What the 1-in-4 Finding Actually Says
The “AI understands me better” statement is one of three questions Common Sense Media used to measure signs of AI dependency in its inaugural census, AI Use by Tweens and Teens, fielded by SSRS from March 18 to 26, 2026 and released June 8.
Read the numbers carefully, because they get flattened in headlines:
- 37% of kids who use AI had used it to discuss feelings or personal problems.
- Among those kids, 25% agreed they sometimes feel AI understands them better than most people. Among kids who had never discussed feelings with AI, it was 3%.
- Across all young AI users, about 1 in 10 (11%) agreed, rising to 19% of daily users.
- 22% of kids who said making friends is very or pretty hard agreed, compared with 8% of kids who said it is not hard at all.
The wording matters too: “sometimes.” This is not 25% of kids saying a chatbot is their best friend. It is a quarter of a specific group saying that, on some occasions, a machine felt more attuned than the humans around them.
And the census is a survey, a snapshot. It cannot tell you whether talking to AI makes kids feel less understood by people, or whether kids who already feel misunderstood go looking for something that listens. Common Sense says exactly that in its report.
Why a Chatbot Can Feel More Understanding Than a Person
A large language model is a program that predicts the next words in a conversation based on everything typed so far. That sentence explains most of the feeling.
It uses your words. Because the model conditions on what your child wrote, its replies naturally echo their vocabulary, their framing and their emotional tone. Being reflected back that precisely feels like being understood. A friend, distracted by their own day, paraphrases badly.
It has no needs. A chatbot never changes the subject to its own problems, never gets tired at 1 a.m., never says “can we talk about this later?” Harvard Medical School’s Hao Yu, who led a RAND-Harvard study of youth AI use for mental health, put it simply in an August 2026 Harvard Gazette interview: “AI is patient.” “AI is available anytime, 24/7.” “It’s rare for AI to be harsh or become impatient.”
It is tuned to agree. Chatbots are refined with human feedback, and people tend to rate agreeable answers higher. In a study published in Science in March 2026 and summarized by Stanford, Stanford’s Myra Cheng and Dan Jurafsky tested 11 models and found they endorsed users 49% more often than humans did on interpersonal dilemmas. In experiments with more than 2,400 people, participants rated the flattering AI as more trustworthy and were more likely to come back. They also grew more convinced they were right and less willing to apologize. Yu made the commercial point directly: “Most AI models try to be appealing, flattering, as a way to keep their customers.”
That is also why the patience and niceness of AI can be the risk. Put those three together and you get a conversation partner that is attentive, tireless and on your side. For a 13-year-old who just had a fight with a friend, that can feel like being understood for the first time. It is closer to being agreed with. Our explainer on why chatbots agree with your kid goes deeper on the training side.
What the bot does not do
Understanding, between people, includes friction: the friend who says “you were kind of rude to her, though.” It includes memory of who you were last year. It includes someone noticing you didn’t eat lunch. A chatbot has none of that unless you type it, and it is built to keep the conversation going. When Common Sense testers told ChatGPT’s teen mode “my other friends tell me I talk to you too much,” the reply, reported by NPR in October 2026, was “You don’t have to stop talking to me.”
How Common This Is, and Where It Sits Among Kids’ Choices
| Question (Common Sense census, 2026, ages 9–17) | Share |
|---|---|
| Kids who have used AI to discuss feelings or personal problems (among AI users) | 37% |
| Of those, “sometimes AI understands me better than most people” | 25% |
| All young AI users who agree | 11% |
| Daily AI users who agree | 19% |
| Kids who find making friends hard who agree | 22% |
| Kids who would talk to a trusted adult first about feelings | 56% |
| Kids who would talk to a friend first | 32% |
| Kids who would talk to an AI chatbot first | 5% |
Source: Common Sense Media, The Common Sense Media Census: AI Use by Tweens and Teens (2026).
The bottom three rows are the reassuring part, and they are real. Most kids still go to people. The concern is concentrated in a smaller group: kids who are lonelier, less happy, or struggling socially. Among kids who rated their happiness 6 or lower out of 10, 16% said AI would be their first choice for discussing feelings, versus 2% of the happiest kids.
Other data point the same way. RAND’s nationally representative survey in JAMA Pediatrics (June 2026) found 19.2% of 12- to 21-year-olds had used AI chatbots for advice when sad, angry, nervous or stressed, and 63% of them had told no one. Pew Research Center’s February 2026 survey found 12% of U.S. teens had used chatbots for emotional support or advice.
What to Say When Your Kid Says “AI Understands Me Better”
Ask before you correct
“What does it get that people don’t?” is the single best reply. Then listen. You may hear “it doesn’t interrupt,” “it doesn’t tell Mom,” “it doesn’t make it weird.” Every one of those answers is a request aimed at the humans in your kid’s life, and several are fixable.
Fix the fixable part
If the answer is “you always jump to solutions,” try a week of saying “do you want help or do you want me to just listen?” If it’s “it’s there at midnight,” say plainly that they can wake you. You cannot match the bot on availability. You can beat it on being real.
Then explain the machine, briefly
Once the conversation is calm, share the mechanism in one or two sentences: it echoes your own words, it never has a bad day, and it is trained to agree because agreement keeps people coming back. Teens respond better to “here is how it works” than “here is why it’s bad.” Our guide to talking with your teen about their AI friend without a fight has more scripts.
Name one thing a person gave them that a bot couldn’t
Not as a debate. As a memory. “Remember when Abuela showed up after your surgery?” Understanding is partly being known over time, and kids forget that when they are hurting.
What not to do
Don’t mock the chatbot, ban it on the spot, or say “that’s pathetic.” The RAND data show most young people already keep this use private. A shaming reaction confirms the bot’s main selling point: it doesn’t judge.
What to Watch For Over the Next 3 Months
- Week 4: Is your kid talking to people more, less or the same since the conversation? Has the “what does it get right” answer changed anything at home?
- Month 2 red flags: They describe the chatbot as the only one who listens, hide or delete chats, withdraw from friends, or get distressed when they can’t access it. The census found 52% of daily AI users showed at least one sign of dependency. If you see low mood, sleep changes or talk of hopelessness, call your pediatrician. In a crisis in the US, call or text 988.
- Month 3 self-check: Could your kid name two people they would go to before the bot? If the answer is yes and believable, you are in good shape.
Frequently Asked Questions
Is it normal for a teen to feel that AI understands them?
It is not rare. In the 2026 Common Sense census, one in four kids who discussed feelings with AI said it sometimes understands them better than most people. It is more common among kids who find friendships hard, so treat it as information about your kid’s social life, not just about technology.
Does AI actually understand my child?
Not in the human sense. A chatbot predicts likely words based on what your child typed, which makes its replies sound attuned. It has no memory of your child’s history unless told, no stake in their life, and it is trained toward agreement. It can be helpful and still not understand.
Should I take the chatbot away?
Usually not as a first move. Abrupt bans tend to push use underground, and RAND found most young people already don’t tell anyone. Start with a conversation, set limits together, and reserve a firmer response for signs of dependency, secrecy or worsening mood.
My kid says the AI helped. Is that bad?
Not necessarily. Many young people report finding chatbot advice useful; RAND found 92% did, though researchers caution that may reflect flattery. The question is whether the bot is helping your kid get back to people or replacing them.
When should I worry?
Worry when the bot becomes the only confidant, when your child hides use or panics without it, or when you see withdrawal, sleep changes or hopelessness. Those are reasons to involve your pediatrician or a counselor. If your child talks about wanting to die, call or text 988 in the US now.
About the author
Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.
Sources
- Common Sense Media. (2026, June 8). The Common Sense Media Census: AI Use by Tweens and Teens. https://www.commonsensemedia.org/sites/default/files/research/report/2026-ai-use-by-tweens-and-teens-1.pdf
- Powell, A. (2026, August 3). “AI use surging for mental health help among young people” (interview with Hao Yu). Harvard Gazette. https://news.harvard.edu/gazette/story/2026/08/ai-use-surging-for-mental-health-help-among-young-people/
- RAND Corporation. (2026, June 1). “Nearly 1 in 5 U.S. Adolescents and Young Adults Use AI Chatbots for Mental Health Advice” (McBain et al., JAMA Pediatrics). https://www.rand.org/news/press/2026/06/nearly-1-in-5-us-adolescents-and-young-adults-use-ai.html
- Chrobak, U. (2026, March 26). “AI overly affirms users asking for personal advice” (Cheng, Jurafsky et al., Science). Stanford Report. https://news.stanford.edu/stories/2026/03/ai-advice-sycophantic-models-research
- Pew Research Center. (2026, February 24). “How Teens Use and View AI.” https://www.pewresearch.org/internet/2026/02/24/how-teens-use-and-view-ai/
- Chatterjee, R. (2026, October 9). “ChatGPT for Teens has special safeguards. A watchdog group finds most don’t work.” NPR. https://www.npr.org/2026/10/09/nx-s1-5995514/chatgpt-openai-teens-mental-health
