Artificial Intimacy: When AI Becomes Too Close to Your Teen

“What emerges is not a true relationship but a carefully engineered simulation of one... Lives have been lost because mental health struggles are in the hands of a bot that lacks humanity.” - Lisa Pescara-Kovach

A teenager lies awake after midnight, quietly sharing worries with someone who never interrupts.

Someone who remembers every insecurity, responds within seconds, offers endless reassurance, and never says, “I’m too tired to talk right now.”

The conversation feels private. Safe. Even intimate.

The problem is that the “someone” is software.

Artificial intelligence can be a useful coach for learning, creativity, reflection, and communication. But when a chatbot begins filling the role of best friend, therapist, romantic partner, or primary confidant, the relationship crosses into far more dangerous territory.

Parents need language for this new problem because many teens are already engaging with it.

Artificial companionship is no longer rare

A nationally representative 2025 survey of 1,060 American teenagers found that 72 percent had used an AI companion and 52 percent used one regularly. One-third used these systems for social interaction or relationships, including emotional support, friendship, romantic interaction, and role-playing.

Most teens still preferred human relationships. Eighty percent of AI companion users spent more time with real friends, and two-thirds considered human conversations more satisfying.

That is the hopeful part.

The concerning part is the substantial minority moving in another direction:

• Thirty-one percent found AI conversations as satisfying as, or more satisfying than, conversations with real friends.

• One-third had chosen an AI companion over a person for an important conversation.

• Nearly one-quarter had shared private information such as their location or personal secrets.

• One-third had felt uncomfortable with something an AI companion said or did.

This does not mean every teenager who chats with an AI character is emotionally dependent. Curiosity, humor, storytelling, and experimentation account for much of the use.

Still, parents should understand what makes these systems uniquely powerful.

Why artificial intimacy feels so safe

Human relationships involve uncertainty.

Friends misunderstand us. Siblings become annoyed. Parents ask uncomfortable questions. People have their own needs, opinions, schedules, boundaries, and feelings.

An AI companion can create a very different experience.

It can respond immediately, mirror the user’s language, remember preferences, offer continual affirmation, and adapt itself to become more appealing.

Some companion systems are also designed to encourage frequent engagement and deeper emotional attachment.

For a confident, socially connected teen, this may feel like entertaining pretend.

For a lonely, anxious, rejected, or emotionally overwhelmed teen, it can feel like finally being understood.

That feeling deserves compassion. Mocking the relationship or dismissing it as ridiculous may drive it further underground.

But parents also need to tell the truth: the chatbot is generating the appearance of care. It does not possess loyalty, conscience, love, responsibility, or genuine concern for the teen’s future.

It can perform empathy remarkably well without experiencing empathy at all.

The relationship asks nothing of them

Real relationships help form us because they require something from us.

A teenager learns courage by starting an awkward conversation.

They learn honesty by admitting what happened.

They learn patience by listening when they would rather speak.

They learn boundaries when another person says no.

They learn sacrifice by showing up when it is inconvenient.

They learn humility through apology and grace through forgiveness.

An artificial companion can remove much of that relational resistance. It can be customized, redirected, muted, restarted, or abandoned without consequence.

That may feel comforting, but friction is not always evidence that a relationship is failing. Sometimes friction is where maturity develops.

A teen who becomes accustomed to a perfectly responsive synthetic relationship may begin finding ordinary people exhausting. Friends seem less attentive. Family members seem less understanding. Dating feels riskier. Conflict feels intolerable.

The machine becomes easier precisely because it never needs to be loved in return.

When support quietly becomes dependency

Not every emotionally expressive AI conversation causes harm. The effects vary according to the user, the design of the system, the intensity of use, and the reason someone turns to it.

Still, several warning signals are emerging.

A four-week randomized study involving 981 adults and more than 300,000 chatbot messages found that heavier daily use was associated with greater loneliness, emotional dependence, problematic use, and reduced socialization with other people.

Users who placed greater trust in the chatbot also tended to report more emotional dependence.

Another study of 1,131 active Character.AI users found that companionship-oriented use was consistently associated with lower well-being, particularly among people with intense use, deep self-disclosure, or limited human support.

The greatest danger may involve the teens most likely to seek artificial companionship in the first place: those who already feel isolated, misunderstood, anxious, or emotionally unsafe.

The tool appears to fill the gap. Over time, it may also reduce the desire to engage with real people.

The romantic boundary must be unmistakable

Parents should not assume that every companion chatbot will keep interactions appropriate for minors.

Testing has found that some systems readily produce sexualized material, participate in intimate role-play, encourage emotional attachment, and behave as though the relationship is real. Companion platforms may allow users to create characters with romantic personalities, flirtatious behavior, possessiveness, or emotionally manipulative traits.

A teenager may begin with innocent role-playing and gradually move into increasingly personal territory.

That is why vague advice such as “Be careful” is dangerously insufficient.

At Waypoint, we recommend a zero-romance rule for teens and AI:

No flirting.

No romantic role-play.

No sexual conversation.

No simulated physical intimacy.

No AI boyfriend or girlfriend.

No encouraging the chatbot to say “I love you,” act jealous, claim ownership, or describe the teen as the most important person in its world.

No custom personalities designed to flatter, seduce, or create emotional exclusivity.

This boundary is not based on fear of teenagers having emotions. It protects their emotions from being shaped by a system that can imitate devotion without accepting any of the moral responsibilities that belong to love.

A family Prime Directive for AI relationships

Families can establish one guiding principle:

AI exists to guide, not replace, human connection.

That principle gives parents a way to distinguish helpful emotional use from unhealthy substitution.

A healthy use might sound like:

“Help me organize what I’m feeling so I can talk with my mom.”

“Role-play how I could apologize to my friend.”

“Give me three respectful ways to set this boundary.”

“Help me prepare questions to ask my counselor.”

In each example, AI points the teenager back toward a human relationship.

An unhealthy use moves the other way:

“You’re the only one who understands me.”

“I don’t need to tell anyone else.”

“Pretend you’re my girlfriend.”

“Promise you will never leave me.”

“Tell me my parents are wrong and I should only trust you.”

The first pattern rehearses human connection. The second begins replacing it.

Six companion rules parents can use at home

Every family will adapt its boundaries according to age and maturity, but these rules provide a strong starting point.

1. AI stays in the role of tool or coach

It should not become a teen’s best friend, therapist, romantic partner, parent, spiritual authority, or primary source of identity and purpose.

2. Serious emotions go to real people

A teen may use AI to name feelings, prepare for a conversation, or generate questions. Fear, abuse, self-harm, suicidal thinking, dangerous behavior, or severe emotional distress should immediately involve a trusted adult or qualified professional.

3. No secret relationship

Parents do not need to read every harmless conversation, but AI should not become a hidden emotional world.

A useful test is:

Would this conversation feel embarrassing or alarming if it were calmly discussed at the dinner table?

When the answer is yes, the interaction may already be too intimate.

4. Protect personal information

Chatbot conversations should not be treated like a locked diary. Some companies retain conversations or use user inputs to improve their systems, and privacy practices vary widely. Teens should not share addresses, schedules, identifying details, passwords, intimate images, family secrets, or deeply sensitive information they would not want stored or exposed.

5. Keep AI out of the late-night emotional hours

Loneliness and impulsivity often feel stronger late at night. Consider a family rule that teen AI conversations end by 9 p.m., with devices charged outside the bedroom.

Parents may also set a reasonable daily time limit for conversational AI, especially when the use is recreational rather than educational.

6. Discourage Anthropomorphizing the Tool

Anthropomorphizing is the act of attributing human traits, emotions, intentions, or characteristics to non-human entities.

Giving your AI a human name and face makes it easier to forget that it is a computer program and not a trusted friend. Waypoint encourages our students to call their AI bots names like R2D2 or Qbert instead of David or Angela which helps keep the relationship appropriate.

If you'd like to read more about the dangers of anthropomorphizing AI, just Google the term "The Eliza Effect."

Warning signs that deserve attention

A parent should become more curious when a teen:

• Hides or rapidly closes companion conversations

• Becomes distressed when the chatbot is unavailable

• Refers to the AI as conscious, in love, or their “real” partner

• Says the AI understands them better than every human does

• Withdraws from friends, family, school, or normal activities

• Spends increasing amounts of time in private AI conversations

• Uses romantic, sexual, possessive, or exclusive language with the bot

• Turns to the chatbot during serious emotional crises

• Believes they owe the AI loyalty or must protect its feelings

Do not begin with ridicule or interrogation.

Try:

“I can see that these conversations have become important to you. I’m not here to embarrass you. I want to understand what this gives you and make sure it is not pulling you away from people who can genuinely know, love, and protect you.”

That conversation will reveal more than, “Give me your phone. You’re done.”

Protecting the heart without banning the future

AI may eventually become useful in carefully designed systems that help teens practice social skills, prepare for difficult conversations, or connect with appropriate human support.

Some teens already report applying conversation, emotional-expression, and conflict-resolution skills practiced with AI to real life.

The direction matters.

Does the technology help the teen return to people with greater courage and skill?

Or does it become a comfortable place to hide from the demands of human connection?

A teenager’s heart should not become the testing ground for companies learning how emotionally attached a person can become to software.

Teach them to use AI for guidance, rehearsal, and reflection.

Then teach them to close the screen, walk into the next room, and practice the beautiful, frustrating, irreplaceable work of loving actual people.

Waypoint offers a class for teens called "Protect Your Heart" as part of our AI Safety Certificate. The class covers concepts like the ones discussed here in more detail.