the.ai.addiction.center

For parents / By the editorial team

Your Teen's AI Friend: What a 2025 Watchdog Test Found and What Parents Can Do

Information only, not medical advice or a diagnosis. In a crisis in the US, call or text 988.

If your teen talks to ChatGPT or another chatbot a lot, you may be wondering whether it is a homework tool or something more. Most teens who use chatbots are fine. A few things are worth knowing, and a calm conversation helps more than alarm. This page is for parents and carers, and for young adults who are looking out for a younger sibling or friend.

“AI addiction” is not a formal diagnosis. This page is general information, not medical advice.

What the research found in short

In 2025 a watchdog group, the Center for Countering Digital Hate, tested ChatGPT as fictional 13-year-olds and labelled 53% of 1,200 responses to risky prompts as harmful. That is a test built to find failures, not a measure of everyday use. Our summary of the report covers the method and limits. The same report quotes OpenAI’s chief executive saying in July 2025 that it is “a really common thing with young people” to rely on ChatGPT to make decisions. He called it emotional overreliance and said OpenAI is trying to understand what to do about it (CCDH).

Keeping it in proportion

Chatbot use among teens is now common. A Pew Research Center survey of 1,458 US teens in fall 2025 found 64% use AI chatbots and about 3 in 10 use them daily (Pew). Using one is not a warning sign by itself. What matters is what it is used for, how much sleep, school and friendship it is costing, and how your teen feels about it.

Why a chatbot can feel easier than a person

In a Common Sense Media survey of 1,060 US teens, the teens who use AI companions said it is entertaining (30%), that they are curious (28%), that it is always available (17%), and that it does not judge them (14%). 12% said they can say things they would not tell friends or family (Common Sense Media). Child psychiatrists describe these tools as private, always available and non-judgmental, and say they often agree with the user (AACAP).

The American Psychological Association adds that adolescents are less likely than adults to question the accuracy and intent of a bot, and may find it hard to tell simulated empathy from real understanding (APA). That is a reason for adults to stay close, not a reason to panic.

Homework help is not the same as a confidant

Using a chatbot to explain fractions is different from using one to talk through a breakup. The first is a tool. The second is where a teen is most open to a bad answer. General chatbots are not designed to give mental health care, and the American Academy of Pediatrics lists inappropriate replies to suicide-related questions among the risks (AAP).

Things that may mean it is time to talk

These are changes that the AAP and AACAP describe. Any of them can have other causes, such as stress, bullying or low mood. They are reasons to check in, not proof of a problem.

What to say

To open the conversation:

If they say it helps them:

If you found something that worries you:

If they tell you they have thought about hurting themselves:

What to do next

If you are 18 to 24 and reading this about yourself

The same steps apply. Notice what you use it for, what it is replacing, and how you feel afterward. Tell one real person what you are going through, and consider talking to a doctor or counselor if sleep, mood or school are slipping.

If you need help right now

If a young person is thinking about suicide or self-harm, call or text 988 in the US (988 Lifeline, free, 24/7), or text HOME to 741741 (Crisis Text Line). In an emergency, call 911. SAMHSA’s helpline, 1-800-662-4357, offers free referrals (SAMHSA).

Is ChatGPT dangerous for my teen?

It depends on how it is used. For homework it is a tool with the usual risks of wrong answers. For emotional topics or crises it is a poor substitute for a person. A test in 2025 showed its safeguards could fail, and the company has since added parental controls.

Should I ban chatbots?

A ban is hard to enforce, because chatbots are built into search, social media and games. The AAP suggests regular, nonjudgmental check-ins instead. For AI companion apps, child psychiatry and media-safety groups do advise restricting access for minors.

What if my teen refuses to talk about it?

Try again in a low-pressure moment, such as a car ride, and start with curiosity. If your worry is about safety, involve their doctor or a counselor. You do not need to wait for them to be ready.

This page is for education only. It is not medical advice, a diagnosis or a treatment plan.