Your Kid Is Confiding in an AI Chatbot. Do You Know What It’s Saying Back?

You can rest easy on one count: AI chatbots probably aren’t bullying your offspring. If anything, AI is often overly agreeable.

“The more common failure is being too accommodating, not putting people down,” said Matthew Freire, managing director of Use.AI. “Models are tuned using human feedback, and people tend to rate agreeable, validating answers more highly, so systems can learn to agree with users, soften disagreement, and mirror how a question is framed.”

But kindness in interactions between AI and children isn’t necessarily a cause for celebration. An excessively compliant AI approach is called sycophancy, and while it may seem harmless, it can create issues for children responding to real-world situations where people disagree with them.

Your child may also come across sexual content through AI. I talked to Teodora Pavkovic, Digital Parenting Expert from Qustodio, a company that provides parental control and online monitoring software. Pavkovic called sexually charged exchanges “a relatively new and unexplored type of online behavior between children and AI bots.” She pointed me to a May 2026 study in the Journal of Adolescence, which found that nearly half of teens who used conversational AI bots reported exposure to one or more specific risks and harms.

A separate July 2026 study in the Journal of Medical Internet Research, which used data from Aura, Qustodio’s parent company, looked at what kids type into AI chatbots. Kids mostly used AI as a tool, like for homework. But about 15% of the time, kids engaged in sexual role-play, and these themes were more common among younger kids.

“The early evidence seems to show that children may sometimes introduce sexual themes in a conversation but are much more likely to experience the AI bot introducing a risky or potentially harmful theme,” Pavkovic explained. Both trends warrant concern about child-AI interactions.

Sycophantic interactions and early exposure to sexual content aren’t the only things to worry about when your kids engage with AI chatbots. Experts I spoke with shared other potential negative outcomes of children confiding in AI.

Kids who frequently confide in chatbots may also experience social displacement, a phenomenon where chatbots push users away from interactions with other people.

“Research in this area is still growing, but one of the early insights has been that people who use chatbots as companions report being lonelier than those who don't,” shared Pavkovic. “The bot not only fails to satisfy their social needs but also doesn't seem to encourage them — at least not successfully — to seek out human relationships.”

Cognitive outsourcing is when a user allows AI to do its thinking for it, and it’s a particularly pervasive problem.

In the past, when kids had an issue they needed solved, they might talk to a parent, do some online research, or even watch a DIY video. Whatever approach they took, they either received encouragement to problem-solve by the parent or put in work on their own by analyzing articles or figuring out video steps.

Now, a child asking an AI chatbot for help may get the answer from a chatbot without much contemplation on the child’s end. As a result, the only thing the kid learns is to turn to the chatbot with their problems.

In the long run, Pavkovic said, this approach “actually leaves us with a deficit in knowledge on the particular subject(s) that we're getting the bot's help with. As a result, we're left feeling less confident as well.”

AI psychosis involves AI-triggered delusions that require professional treatment, including but not limited to believing an AI is a true sentient being. AI psychosis is rare, but it’s always possible, even in individuals who have no history of mental health crises.

“In some of the more extreme cases, we've seen that longer interactions with an AI bot can lead people to believe that they are speaking with a sentient being, at times resulting in the development of serious mental health issues known as ‘AI psychosis.’” Pavkovic said.

How do you know if your child engages too much with or becomes too dependent on AI chatbots? Here are some signs to look out for, provided by Brittany Caputo, a licensed behavior specialist and behavior consultant in Pennsylvania:

• Increasingly choosing conversations with AI over interactions with friends or family
• Becoming unusually distressed when access to a chatbot is interrupted
• Relying on AI as their primary source of emotional reassurance or decision-making
• Withdrawing from previously preferred activities
• Changes in sleep, mood, or school participation associated with chatbot use

“These behaviors don’t automatically indicate dependency, but patterns of change are worth paying attention to,” said Caputo.

Dr. Christine Ghilain, a double board-certified clinical and pediatric neuropsychologist in New Jersey, commented that symptoms of AI chatbot dependency can “overlap with other things, like anxiety, depression, attention differences, or social struggles. A child who leans on a chatbot is often telling us that people and relationships feel hard right now.”

You may (understandably) just want to keep your child away from AI chatbots altogether. Sadly, it’s not fully up to you. While forbidding certain recreational chatbots or chatbot activities is a reasonable boundary for your home, kids these days may still require access to AI for school assignments and, later, jobs.

However you approach the AI problem, you can set up your child with healthy boundaries for the future by using these guidelines.