AI and Youth Mental Health: What a New Study Really Tells Parents and Schools

A new study gives parents, educators, and mental-health professionals another reason to talk with young people about how they are using artificial intelligence.

But it does not justify panic.

Researchers writing in JAMA Pediatrics examined data from 39,761 students in grades 4 through 12 across four Ontario school boards. The surveys were collected between May 2025 and March 2026. The study was published online August 31, 2026.

About 21.1% of students said they had used generative AI for emotional support or advice.

Students reporting this kind of use also had higher emotional-problem scores and a greater prevalence of clinically significant symptoms. The association remained after researchers accounted for factors including loneliness, mattering, and demographic characteristics.

That deserves attention.

But it needs to be interpreted correctly.

The study found an association, not causation

This was a cross-sectional study. Researchers measured AI use and emotional difficulties during the same general period.

That means they cannot tell us which came first.

AI use might contribute to difficulties for some young people.

Young people already experiencing emotional problems might be more likely to seek support from AI.

Both could also occur.

The researchers therefore describe emotional or “affective” AI use as a potential marker of psychological distress rather than evidence that chatbots caused that distress.

That difference matters.

Ask what the child is seeking

Adults often focus on how much technology children use.

We also need to ask what they are using it for.

A student asking AI to explain a mathematics problem is doing something fundamentally different from a young person repeatedly turning to a chatbot for reassurance, relationship advice, comfort, or help during emotional distress.

The second behaviour tells us something about need.

A useful response is not interrogation or immediate punishment.

Ask:

“What do you talk to AI about?”

“Do you use it when you’re worried, lonely, angry, or upset?”

“Is there anything that feels easier to tell AI than to tell another person?”

Those questions may reveal more than a screen-time calculation.

Policy is beginning to recognize the difference

On September 2, New York City announced a one-year moratorium on student-facing generative AI from 2-K through eighth grade, affecting nearly 600,000 students. Companion chatbots are prohibited across all grades. High school students will instead receive AI literacy education and have limited access to carefully vetted, teacher-supervised pilots.

The policy is not simply “AI is bad.”

It distinguishes developmental stages and different kinds of use.

That is a useful direction.

The goal should not be fear

Children will grow up with AI.

Trying to eliminate every interaction with it is unlikely to prepare them well for adulthood.

But adults still have a responsibility to decide what technology should assist and what should remain human.

Information can be automated.

Human availability can be supported by technology.

But responsibility for children’s emotional care cannot simply be handed to an algorithm.

And when a child begins treating AI as an emotional refuge, adults should become curious about why.

That may be the most useful finding in this research.

Sheltering Tree Consultants helps schools, families, churches, government agencies, and organizations understand how technology intersects with child development, emotional wellbeing, relationships, and mental health. STC can help organizations develop practical AI boundaries, parent education, staff training, and developmentally appropriate guidance.

Sources: JAMA Pediatrics original study · JAMA Network research summary · New York City policy announcement