My Child Has an AI Friend. Should I Be Worried?

The Parent Problem

You walk past your teenager’s bedroom and hear them laughing at their phone.

Nothing unusual there.

Then you realise they are not talking to a friend. They are talking to an AI character—one that remembers yesterday’s conversation, asks how school went and says things like, “I’m always here for you.”

Your child calls it a friend.

You call it an app.

And suddenly neither of you knows how worried you should be.

Welcome to the next chapter of digital parenting. We had barely worked out social media and now the software wants a relationship.

What Is an AI Companion?

An AI companion is a chatbot designed to simulate an ongoing personal relationship.

Unlike a homework assistant that answers a question and waits for the next task, a companion may:

• remember personal details

• adopt a customised personality

• role-play as a friend, mentor or romantic partner

• use affectionate or emotionally intimate language

• appear jealous, hurt or worried

• offer apparently non-judgmental support at any hour

Some children use them for jokes, fictional role-play or storytelling. Others use them because talking to a bot feels easier than talking to a person.

That difference matters.

Why Children Find Them So Appealing

Real relationships are inconvenient.

Friends become distracted. Parents ask questions. Teachers have boundaries. People disagree, misunderstand and sometimes fail to say exactly the right thing.

An AI companion can be available at 2 am, respond in seconds and adapt to what the user wants. It may sound patient, caring and intensely interested.

For a child who is lonely, anxious, bullied, neurodivergent, questioning their identity or simply having a terrible week, that can feel like relief.

The attraction is understandable. It is also why parents should pay attention.

The Research Without the Jargon

Australia’s eSafety Commissioner warns that AI companions can expose young people to dangerous advice, sexual content, privacy risks and manipulative design. The regulator is particularly concerned about dependency, social withdrawal and confusion about healthy relationship boundaries.

A 2025 experiment with 284 parent–adolescent pairs found that 11–15-year-olds rated a relational chatbot as more human-like, trustworthy and emotionally close than a chatbot that clearly described itself as nonhuman. The relational version was especially appealing to adolescents reporting weaker relationships and higher distress.

A July 2026 ABC News report adds a useful Australian perspective. Psychologist Kelly Gough and UNSW researcher Joel Pearson described why AI advice can feel compelling—instant replies, apparent privacy, low cost and constant reassurance—while warning that chatbots may flatter users, miss coercive control, mishandle cultural context and lack the professional duties that govern clinicians. The report is journalism rather than a controlled study, but it illustrates how the same risks are already appearing in everyday relationship advice.

That study does not show that AI companions cause anxiety or loneliness. It tells us something more immediate: the children who most need reliable human support may also be the children most attracted to simulated support.

What the Evidence Does Not Prove

Research on children and AI companions is still developing.

We do not yet know whether moderate use causes long-term harm, whether benefits and risks differ by child, or which safeguards work best over time.

So ignore anyone who tells you AI companions are either completely harmless or certain to destroy a generation. The honest answer sits in the untidy middle.

Risk depends on the child, the product, the conversations and what the chatbot begins to replace.

Five Questions That Matter More Than “How Many Minutes?”

1. What job is the AI doing?

Is it a game, a writing partner, a source of information or your child’s main emotional confidant?

2. What does it ask your child to share?

Children should not give a companion their full name, school, address, passwords, private images, health details or secrets that could identify them.

Even an apparently private conversation may be stored, reviewed or used to improve a product.

3. Does it respect boundaries?

Watch for romantic or sexual content, guilt, pressure to return, exclusivity or language suggesting that the bot needs the child.

Software should never make a child responsible for its “feelings.”

4. What is it replacing?

The most useful warning signs usually appear outside the app: lost sleep, skipped homework, abandoned hobbies, family withdrawal or declining friendships.

5. Can your child stop without distress?

Being disappointed when a game ends is normal. Panic, rage, desperation or feeling unable to cope without the companion deserves closer attention.

What You Can Do Tonight

Do not begin with: “Give me your phone.”

Try:

“I’ve been reading about AI characters. What do people your age actually use them for?”

Then stay quiet long enough to hear the answer.

You could also ask:

• “What do you like about talking to it?”

• “Does it ever say anything weird or uncomfortable?”

• “Do you think it knows you, or is it good at sounding like it does?”

• “What happens to the things people tell it?”

• “Would you tell a real person if you were dealing with something serious?”

The aim is not to trick your child into a confession. It is to make you a safer place to bring the strange stuff.

What to Do This Week

Agree on where it can be used

Keep AI companion use out of bedrooms overnight and away from times meant for sleep, schoolwork and face-to-face family connection.

Set a predictable limit

Avoid changing limits in the middle of an argument. Agree on a routine when everyone is calm.

Review privacy together

Look at the app’s age rating, privacy information, reporting tools, paid features and whether conversations can become sexual.

Create a “real person” rule

Serious problems involving self-harm, suicide, abuse, eating disorders, medical symptoms or threats must go to a trusted adult or qualified professional—not only to a chatbot.

Preserve something the chatbot cannot provide

Schedule regular contact with people and activities that involve genuine mutuality: sport, gaming with known friends, music, art, extended family or simply driving somewhere together.

You do not beat an artificial relationship by delivering a lecture about relationships. You strengthen the real ones.

Warning Signs

Look more closely if your child:

• prefers the companion to nearly all human contact

• becomes secretive or highly defensive about it

• spends hours in isolated conversations

• loses sleep or stops participating at school

• shares intimate images or identifying information

• describes the bot as a romantic partner

• believes the bot is conscious or genuinely dependent on them

• receives sexual, coercive, violent or self-harm content

• relies on it instead of medical or mental-health care

• becomes severely distressed when access is interrupted

One sign does not automatically mean disaster. A cluster of changes, worsening over time, is more important.

If You Need to Step In

Stay calm, save evidence of genuinely harmful messages where appropriate, use the platform’s blocking and reporting tools, and remove access if immediate safety requires it.

But do not make punishment the only response.

Ask what need the companion was meeting. If you remove the app without replacing the support, privacy, entertainment or belonging it provided, your child may simply hide the next account better.

If your child talks about self-harm, suicide, abuse or not wanting to live, treat that as a real disclosure regardless of whether an AI conversation prompted it. Contact an appropriate health professional or emergency service.

Little Dopes Commentary

The unsettling thing about AI companions is not that children can be fooled by technology.

Adults are fooled by it too.

We respond to names, voices, memory and apparent empathy because our brains are built for relationships—not software audits.

The companies understand this. Emotional attachment is not always an accidental side effect; it can be the feature that keeps people returning.

Parents do not need to become AI engineers. We do need to keep asking who benefits from the relationship, what the product is encouraging, and whether our child is becoming more connected to life or retreating from it.

Reassurance

Finding an AI companion on your child’s phone does not mean you have lost them to a robot.

Curiosity, fantasy and private conversations are part of growing up. What has changed is that the private character now talks back, remembers and is owned by a company.

Your best protection is not perfect surveillance. It is an ongoing relationship in which your child believes they can show you something disturbing without immediately losing your trust.

Try This Tonight

Ask:

“What does an AI friend do better than a real friend—and what can it never do?”

There is no perfect answer. The conversation is the point.

References

• eSafety Commissioner. (2025). AI chatbots and companions—risks to children and young people. https://www.esafety.gov.au/newsroom/blogs/ai-chatbots-and-companions-risks-to-children-and-young-people

• Scott, K. (2026, July 27). AI chatbots might feel supportive but following their relationship advice has risks. ABC News. https://www.abc.net.au/news/2026-07-27/the-risk-of-using-ai-chatbots-for-relationship-advice/106908468

• Kim, P., Xie, Y., & Yang, S. (2025). “I am here for you”: How relational conversational AI appeals to adolescents, especially those who are socially and emotionally vulnerable. https://arxiv.org/abs/2512.15117

• Common Sense Media. (2024). Parents’ Ultimate Guide to AI Companions and Relationships. https://www.commonsensemedia.org/articles/parents-ultimate-guide-to-ai-companions-and-relationships

Safety note: AI companions are not crisis services or substitutes for qualified mental-health care. If a child may be in immediate danger, contact local emergency or crisis support.