Who Is Speaking to Your Teen? Raising Discernment in an AI World
One of the strange things about parenting teenagers is that we spend years teaching them to listen to us, and then they reach an age when almost every other voice sounds more interesting. A friend says something, and it carries weight. An influencer says it, and suddenly it is worth considering. Then artificial intelligence arrives with an additional advantage: it is available at two in the morning, and it is extraordinarily difficult to annoy. I can type something into AI that is mediocre and somehow receive a response that makes me feel like I may have just solved Western civilization. Imagine how appealing that can be to a teenager still figuring out who they are.
The podcast conversation looked at new teen focused ChatGPT protections designed to make the distinction between AI and human relationships clearer, including reminders that teens are interacting with a chatbot and additional protections for high risk conversations. Those sound like meaningful attempts to address real concerns. But Dr. Kathy made an important point: even strong technological safeguards cannot do the work parents were designed to do. A product can limit what it says. It cannot secure a child’s heart.
Teens Need People Who Will Sometimes Say No
This may be one of the biggest differences between a chatbot and a loving parent. AI is designed to be responsive and helpful. Parents have to be willing to be inconvenient. Sometimes love sounds like, “No.” Sometimes it means pointing out that a choice does not align with family values or that a path is likely to cause pain later.
Teenagers do not always experience those moments as loving. That is partly because they are still developing the ability to evaluate cause and effect and interpret correction without turning it into a statement about their whole identity. Dr. Kathy’s point was that young people need practice learning that correction does not automatically mean rejection. Dad saying, “That was a bad decision,” is not the same as Dad saying, “You are bad.” A chatbot that endlessly affirms can unintentionally make real relationships feel harsher simply because real relationships sometimes require truth.
That is why Dr. Kathy talks about parents needing both a red pen and a green pen. We should absolutely notice what is going well. Celebrate progress and tell kids what they are doing right. But love also corrects. Criticism simply points out what is wrong; correction helps put something right. If our children only experience affirmation, they may begin to confuse encouragement with truth and disagreement with hostility.
Guardrails Cannot Replace a Secured Heart
In Start with the Heart, Dr. Kathy emphasizes the importance of securing a child’s heart so that parents can remain influential. We do that by looking for progress rather than demanding perfection, correcting rather than criticizing, celebrating what is good, and helping children interpret comparisons before they swallow their identity. Those are profoundly human tasks. They require knowing the child standing in front of us.
AI can know what a teenager typed. It may even remember patterns from previous conversations. But a parent knows the face that appears after a difficult day. We know the friend whose name suddenly stopped being mentioned. We know whether silence means anger, embarrassment, or “I will tell you in forty five minutes if you stop asking me questions.” That context is part of a relationship. It is also part of protection.
The more our children trust that we love them enough to tell the truth, the less likely they are to believe that every endlessly affirming voice deserves authority. That does not mean they will always appreciate our involvement. Dr. Kathy was clear that parents need to be “in their business.” That can produce pushback. It is still our job.
AI Can Help With Learning Without Becoming the Teacher of the Heart
The conversation also looked at AI tools designed to support learning by asking questions and providing additional context rather than simply supplying an answer. There are situations where that could be useful, particularly for students who are overwhelmed or need another way into difficult material. Dr. Kathy specifically noted that a student struggling with dyslexia or fear around learning might benefit from tools that lower the initial barrier.
But usefulness is not the same as formation. The danger is that teenagers may become very good at obtaining information without learning how to wrestle with it. Information is not wisdom. Knowing the date of a war is not the same as understanding why it happened, what people believed, what followed, and whether the account you received is even trustworthy.
That is where struggle has value. I do not particularly enjoy reading the manual in my glove box when I am trying to figure out why something in my car stopped working. I would much rather receive a clean three step answer. But when I have had to trace the problem and discover why the part works the way it does, I know the car differently afterward. Sometimes the frustrating parts of learning create the context that makes knowledge durable.
Do Not Let Convenience Remove the Work of Discernment
This may be one of the greatest educational challenges AI creates. The answer comes before the student has had enough time to develop the question. The context can be summarized before the teenager has learned how to locate context. A patient chatbot can guide them step by step, but eventually the child still has to ask, “Is this true?”
Parents and teachers should preserve that muscle. If AI provides an explanation, check another source. Open the textbook. Look at the primary document. Ask the teacher. Ask Dad. Compare accounts. Teach children that receiving an articulate answer is not the same thing as receiving a trustworthy one.
The goal of education is not simply acquiring enough information to finish the assignment. We want young people who can use knowledge wisely, place facts inside larger realities, understand people, recognize competing claims, and make judgments. AI can contribute information to that process. It cannot be allowed to become the process.
The Serpent Was a Conversationalist
Genesis 3 gives us a remarkably relevant picture. The serpent does not appear by smashing through the garden and announcing that he is evil. He starts a conversation: “Did God actually say...?” He positions himself as someone who understands the situation and perhaps understands Eve’s interests better than God does.
The danger is relational before it becomes behavioral. A voice presents itself as trustworthy and begins reframing reality.
That is what struck me as we talked about teenagers interacting with AI. The concern is not that a chatbot is literally the serpent in Genesis. It is that our children are entering a world filled with conversational voices that can sound knowledgeable, intimate, affirming, and certain without deserving their trust. The question they need is not simply, “What did this voice say?” It is, “Who is speaking to me, and why should I trust it?”
That question matters when the voice belongs to AI, a TikTok personality, a friend, a teacher, a political figure, an advertiser, or even a pastor. Discernment asks what authority the speaker has, whether the claim aligns with the truth, what evidence supports it, and what the voice invites us to become.
Our goal cannot simply be silencing every questionable voice. We will lose that battle. The world our children are entering will contain more voices and more personalized influence than anything we experienced growing up. We have to form people capable of hearing a voice and remaining anchored enough to evaluate it.
So use AI when it is genuinely useful. Let it help explain something difficult, or open a door into curiosity. But do not surrender the slower human work. Talk with your teen. Correct them. Celebrate them. Read things together. Let them watch you change your mind when evidence requires it. Teach them to cross reference. Give them space to struggle with something before rushing to the answer. Most importantly, keep directing them toward the voice that deserves ultimate trust. The safest teenager is not simply the one protected by better technological guardrails. It is the teenager who increasingly knows how to say, “I know who God is. I know what is true. Now I can decide whether this voice deserves to lead me.”
Engaging AI and Discernment Through the 8 Great Smarts
Word Smart: Have your teen compare an AI answer with a book, article, Scripture passage, or conversation with a knowledgeable person. Ask them to explain the differences in their own words.
Logic Smart: Teach them to identify assumptions, evidence, cause and effect, and possible errors in an AI response. Ask, “How would we know whether this is actually true?”
Picture Smart: Invite them to diagram where information came from or visually map the larger context around an answer. What is missing from the picture the AI presented?
Music Smart: Discuss how tone and repetition influence trust. Compare the emotional impact of a persuasive song, speaker, advertisement, or chatbot with the truthfulness of its message.
Body Smart: Move learning away from the device when possible. Build, repair, experiment, practice, visit, or physically experience something the technology explained.
Nature Smart: Use AI to generate a question about the natural world, then go outside and test, observe, compare, or investigate the answer in creation.
People Smart: Keep people in the learning loop. Encourage your child to take an AI generated answer to a teacher, parent, mentor, sibling, or expert and ask what they think.
Self Smart: Ask your teen to notice how AI makes them feel. Do they feel unusually affirmed, dependent, understood, frustrated, or tempted to keep asking until they receive the answer they want?
Remember: However your child is smart, give them this habit: before trusting the answer, ask who is speaking, what they want you to believe, and whether that voice deserves authority.