In March 2026, researchers surveyed 1,204 American kids between the ages of 9 and 17 about artificial intelligence. Eighty-six percent said they use it. Nearly a quarter said they use it every day.
Then the researchers asked the kids who use AI what they had asked it. Fifty-seven percent said they had asked it for information or advice about their health or their body. Thirty-seven percent said they had used it to talk about their feelings or personal problems.
It rises with age. Among kids who use AI, 42 percent of the 9 to 12 year olds had asked about their health or body, 60 percent of the 13 to 15 year olds, and 75 percent of the 16 and 17 year olds. These figures describe kids who use AI, which is most kids. They still describe a lot of children asking a machine questions they may not have asked a person.
Before anyone panics, a few things are worth knowing.
Being raised in the 1950s and 1960s, we were typical children and teenagers. Hormones raging, emotions constantly vacillating, and too embarrassed to ask for help with the inner turmoil. Not much has changed.
What has changed is where the questions go. Every generation finds somewhere to take what it is too embarrassed to ask out loud. For some of us it was an older cousin, a worn health book at the library, or a friend who claimed to know. The questions are old. The place is new.
What makes this place different is how easy it is. A chatbot answers at any hour and has no face to look shocked. A child can ask the third version of an embarrassing question without anyone watching their reaction. The appeal isn't hard to see.
A chatbot can give an answer. It can't see your child's face, and it doesn't know your daughter has been quieter all week. Its answer to a health question is not a doctor's answer.
Two more findings deserve a careful read.
Kids who use AI every day or every week were more likely to say they feel lonely at least some of the time. Fifty-four percent of daily users said so, compared with 33 percent of kids who use AI less than monthly or not at all. The researchers are careful here. This does not show that AI makes kids lonely. A lonely kid may reach for a chatbot. Heavy use may deepen loneliness. Both could be true. One survey cannot sort that out, and neither can I.
Eleven percent of kids who use AI said it sometimes understands them better than most people do. Among daily users it was 19 percent. I won't tell you what that means. I will tell you it is a sentence you would want to hear from your child, not from a survey.
One more number. About one in six kids who use AI chatbots said they had seen something inappropriate. Of the kids who had, a third told a trusted adult. More than half did not.
More than 4 in 10 kids say no parent or guardian has ever talked with them about AI safety. That is not a verdict on parents. Nobody handed us a manual. But it does mean that for many kids, the chatbot got there first.
In the last piece in this series we talked about the homework side of this. This is the other side of the same conversation. A homework habit can be turned around. A child who has stopped bringing hard questions to you is a longer road.
You don't need to take the device away, and you don't need to become an expert. You need to stay the person your child would rather ask.
A chatbot can be a fine place to look something up. It is a poor place to be known. You are still the one who can do that.
Source: Common Sense Media, "Census: AI Use by Tweens and Teens," released June 2026. A national survey of 1,204 U.S. children ages 9 to 17, conducted March 18 to 26, 2026, margin of error plus or minus 3.8 percentage points. The loneliness finding is an association, not proof of cause.
This piece was written by Dave Edwards in collaboration with an AI assistant.
Coming soon to this post.