A teenager sends her mother a text.
Mom, I’m so annoyed with my friend right now. I don’t even know what to say to her.
Her mother reads the message, thinks for a moment, and—wanting to be helpful—opens Claude. She pastes in the situation.
The response is thoughtful. It is measured. It identifies the teenager’s feelings, considers the friend’s perspective, suggests a few possible responses, and recommends waiting until emotions have settled before addressing the conflict.
And her daughter replies: Mom. I wanted to talk to YOU.
That exchange captures something important about the emerging relationship between children, parents, and artificial intelligence. The problem isn't necessarily that AI advice is wrong. The problem is that accuracy isn't the same thing as intimacy.
Parents are navigating an increasingly strange world. Their children have access to technologies that can instantly explain a conflict, interpret a text message, generate a response, provide emotional validation, brainstorm solutions, or offer a seemingly endless supply of information. And parents have access to those same tools. Naturally, many parents are tempted to use them.
If a child says, “I think my teacher hates me,” why not ask AI for a thoughtful way to respond?
If a teenager is struggling with a friend, why not get some perspective?
If a young adult sends a long, complicated text about a breakup, why not use AI to make sure your response is supportive rather than reactive?
There is nothing inherently wrong with any of this.
In fact, AI can sometimes help parents slow down. It can help them find words when they are overwhelmed. It can introduce perspectives they had not considered. It can even prevent a parent from immediately responding with something like, “Well, if you had just listened to me in the first place…”
But there is a subtle danger when AI becomes the invisible third person in every parent-child conversation.
A child may not actually be asking:
“What is the objectively best response to this situation?”
They may be asking:
“What does my mom think?”
Those are very different questions. There is something uniquely reassuring about having someone who knows your history respond to you. A mother may remember that her daughter has always had difficulty feeling left out. She may remember the friend from middle school who broke her heart. She may know that her daughter tends to catastrophize when she is tired. She may know that underneath the anger is usually hurt. She may also know when her daughter doesn't actually want advice. She just wants someone to say:
“Oh honey. I know exactly why that hurt.”
That response might not be the most sophisticated psychological intervention. It might not identify every cognitive distortion. It might not provide five communication strategies and a beautifully worded text message. But it carries something AI cannot manufacture simply by analyzing the contents of a conversation: shared history.Mom knows you. Not just the situation. And you know Mom, and what to expect from Mom, and having that expectation met is also comforting. Her thoughts and words specifically are what might heal just because they're hers, not because of what's being said.
There is a growing phenomenon among young people who are becoming accustomed to AI as a conversational presence: they are learning to recognize the language of artificial emotional support. They know what the carefully structured response sounds like. “It makes sense that you feel that way.” “There are several possibilities to consider.” “You might want to communicate your needs directly.” “Before responding, give yourself some time to process.” These statements can be genuinely helpful. But imagine hearing them from your parent after sending a vulnerable text. You might think:
Did you really say that because you know me? Or did you ask an AI what to say?
That distinction matters. Children and adolescents spend years learning whether their parents are emotionally available to them. They learn whether Mom listens, whether Dad gets defensive, whether their parents remember what matters to them, whether their feelings are tolerated, and whether difficult conversations can happen without immediately becoming lectures. A parent-child relationship isn't built exclusively through good advice. It is built through being known.
There is a difference between helping and outsourcing
This doesn't mean parents should never use AI. There are wonderful ways parents can use these tools. A parent might ask AI: How can I respond without becoming defensive? What might my child be trying to communicate that I'm missing? Can you give me three ways to ask about this without turning it into a lecture? That kind of use can actually make a parent more emotionally available, not less. The problem arises when the parent stops doing the emotional work themselves, and becomes a copy-paste machine. The child can recognize when it's not the parent in all their honest truth.
Instead of thinking, What does my child need from me right now?, the question becomes, What does the AI recommend I say? Those aren't interchangeable. One is a relationship. The other is a strategy. And children can feel the difference. Sometimes the best response is truly imperfect. Maybe Mom tells her daughter, “I think you're taking this too personally,” when what her daughter really needs is validation. Maybe Dad launches into a story about his own teenage years that has absolutely nothing to do with the situation. What if she just says: “I don't know what to tell you. That really sucks.” And then sits beside her daughter? That may be enough, it may be even better. Because sometimes the point of texting a parent isn't to obtain a solution. It is to reach across the distance between two people and say: Are you there?
As AI becomes increasingly integrated into family life, parents don't need to reject it. They may simply need to remember what it is—and what it isn't. AI can be a sounding board. It can be a source of information. It can help us find language, consider possibilities, organize our thoughts, and even recognize patterns we might otherwise miss. But it doesn't have to become the voice between parents and children.
Parents can use AI privately to become better listeners. They can ask it to challenge their assumptions or help them approach difficult conversations with greater patience. Then they can close the app. And answer their child themselves.
Because when a child says, “I texted you,” what they may really mean is:
“I wanted to know what you think. I wanted to feel known by you. I wanted my mom—not an algorithm—to answer.”
Explore more of my work on adolescents and AI: