“Daddy, can I speak with ChatGPT now?”

It is easy to imagine this becoming an increasingly familiar question in many homes.

“We’re all busy, and surely the additional language exposure will be helpful. I mean, I have a million things to do right now…” a parent might think.

“Just for half an hour.”

But are there any long-term risks when young people regularly converse with large language models?

ChatGPT’s voice mode, for instance, seems technically impressive. OpenAI has also recently launched ChatGPT for Teens.
The Guardian, 18 August 2026:

On first reading, stronger safeguards, parental controls and learning support seem like positive steps. The new experience is aimed at 13–17-year-olds, although younger children may also encounter AI voice systems informally in shared family settings.

But I’m increasingly questioning what else is being learned—and what isn’t.
The turn-taking feels one-sided. ChatGPT is endlessly patient and often overly affirming, sometimes to the point of sounding sycophantic: “That’s correct.” “Well done.”

Human conversation is messier. Children must listen, wait, read tone, share attention, repair misunderstandings and cope with peers who may be distracted, impatient, jealous, blunt or rude.

Could AI conversations be building vocabulary and confidence while under-exercising listening, empathy and resilience? Could some children accustomed to a permanently supportive conversational partner experience a “social shock” when interacting with real peers?

What effect might that have? Does anyone actually know?

We have introduced an intelligent “thing” into children’s environments without really understanding its long-term effects.

Therefore, alongside safeguards, I would like to see independent, long-term research covering different age groups. Families also need privacy-conscious insights into usage, topic areas, tutoring versus answer-giving, academic shortcutting and overdependence.

Most importantly: to what degree are AI conversations displacing human ones?

Age-appropriate safeguards are a good first step. The harder question is: what kind of human communicator is this technology helping a young person become

How are other parents and educators finding the balance?