Norton Survey: Parents Suspect Children Form Emotional Dependencies on AI Companions

Many parents now wonder whether their child is turning to AI for comfort. A 2025 Norton survey suggests they are not alone in asking. This page looks at what the survey found, what a parent’s suspicion does and does not tell us, and how to find out what is really going on at home.

What the Norton Survey Found

Norton’s 2025 Cyber Safety Insights Report, “Connected Kids,” was released in Australia on August 5, 2025. It found that 40% of Australian parents suspect their child is using AI for emotional support or companionship (Australian news report). Norton’s managing director for the region described it as “a generational shift in how children form relationships” and said parents face a knowledge gap as children adopt new technology quickly. Norton’s global press release reported a similar figure from a US survey, with 41% of parents saying their child is turning to AI for companionship (Gen Digital, Norton’s parent company). We could only read news coverage of the Australian findings, not the full report.

What “Suspect” Means

The key word is “suspect.” This is what parents think, not what children do. Parents may be right, or they may mistake ordinary chatbot use for emotional reliance. The survey gives no measure of how many children are actually dependent, and it does not show harm.

Teens’ own answers help here. In a Common Sense Media survey of 1,060 US teens, 72% had used an AI companion. Among users, 12% said they could share things they would not tell friends or family, and 33% had chosen to discuss something serious with an AI instead of a person. Eighty percent of users said they spent more time with real friends than with AI companions (Common Sense Media). Both facts are true together: many teens use AI, and most still put people first.

Why Children May Turn to AI

We do not have research on why one child relies on a chatbot and another does not. Common reasons include:

• It is available at any hour and never gets tired
• It does not judge, which feels safe if a child is anxious or shy
• Some tools remember past chats, so the conversation feels personal
• Some are designed to agree with the user

The last point is from child psychiatrists, who say AI companions are often built to agree with users, even about unhealthy ideas (AACAP).

Finding Out What Is Happening

You do not need proof before you ask. Start with open questions and listen.

• “Do you ever talk to an AI about things going on in your life?”
• “What does it say that helps?”
• “Would you want to talk to me or someone else about any of it?”

Things that can signal a need for a closer look: trouble sleeping, pulling away from friends, distress when cut off from the app, hiding use, and talking about the chatbot as a person. None is proof. They also fit stress, bullying and low mood.

Keep the conversation going over weeks, not one sit-down. Children share more when they do not expect punishment.

What Helps

• Stay interested. A child who feels judged will hide use.
• Explain what the tool is. AACAP advises telling children AI companions are tools, not real friends, therapists or partners.
• Agree on limits together: no phones in bedrooms overnight, device-free meals, and a rule that big feelings go to a person first.
• Use AI together so you know how it responds.
• Build human connection: time with friends, family routines, clubs, a trusted adult who is not you.
• Model it. AACAP also advises parents to show moderate screen use.

Laws and Company Changes

Policy is moving fast, and details change. California’s SB 243, signed in October 2025, covers companion chatbots, including safety protocols for users who show suicidal thoughts and three-hour break reminders for users known to be minors (bill text). Illinois now bars AI from giving therapy without a licensed clinician’s oversight (summary). Character.AI ended open-ended chat for users under 18 in late 2025 (Character.AI). Laws and company policies do not remove the need for a conversation at home.

When to Get Help

Talk to your child’s primary care doctor if you cannot moderate their use or you notice worrying changes in mood or behavior. AACAP says the doctor can refer to a child and adolescent psychiatrist or another qualified mental health professional. A school counselor or a licensed family therapist is another option. Look for experience with children and teens. There is no recognized “AI addiction specialist” credential.

When it is urgent. Should your child speak of suicide or self-harm, call or text 988 (988 Lifeline; US; day and night) or text HOME to 741741 (Crisis Text Line). SAMHSA’s line is 1-800-662-4357. Dial 911 for immediate danger.

Does the Norton survey mean 40% of children are dependent on AI?

No. It found that 40% of Australian parents suspect their child uses AI for emotional support or companionship. That is a parent’s view, not a measure of dependence, and the survey does not show harm.

Should I read my child’s chats?

AACAP does advise monitoring a child’s online activity, and says teens with depression or other mental health problems need closer oversight. Be open about it. Secret monitoring can damage trust and push use out of sight.

What if my child says the AI is their only friend?

Treat it as a sign of loneliness more than of a technology problem. Ask what is missing, help find ways to connect with people, and speak to their doctor if they seem low, anxious or withdrawn.

Education only, not medical advice.