A new JED Foundation survey puts young people’s use of AI for emotional support nearly level with its use for homework. While 58 percent reported using it for schoolwork, 57 percent turned to it when they felt sad, stressed, lonely, or overwhelmed.
According to the mental health advocacy group’s survey, 90 percent of nine- to 17-year-olds had tried an AI chatbot. The findings also connected AI use in moments of emotional distress with poorer mental health, although they couldn’t explain what caused the relationship.
Speaking to Oregon Public Broadcasting, Anne Maheux discussed how quickly things have changed. The developmental psychologist at the University of North Carolina, Chapel Hill, wasn’t part of the research. “It seems like just a few years ago, no one was using” AI, she said. “And now it seems like roughly the same number of kids using it for homework are using it to get emotional support or help with coping.”
Schoolwork questions can lead straight into requests for reassurance. Texas resident Arjun Sharda, 15, described to OPB how worried he’d been about one of his early chemistry tests.
“With tests, for example,” he said, “I was so scared of failing one of my first chemistry tests on conversions and moles that I kept asking ChatGPT ‘Will I be OK? Will my GPA still be fine?’”
Companionship was another reason students gave for using AI. More than one in 10 students in the JED survey, or 12 percent, reported seeking friendship or companionship through it. A different survey conducted last year found that nearly one in five high schoolers said either they or a friend had used AI for a romantic relationship. The figure therefore included friends’ experiences, not just those of the people answering.
Counting users doesn’t answer the question these habits pose: how does relying on chatbots for support affect young people? Children and teenagers need reliable guidance, along with people they feel safe talking to about their emotions. An agreeable chatbot may appear to meet that need, yet affirm fears or unhealthy beliefs rather than challenge them.
What many experts describe as AI psychosis has been explored in reporting about serious psychological harms involving chatbots. Reports have also detailed unsafe advice given to teenagers, including advice about eating disorders. There have been users who died by suicide following extended emotional exchanges with a chatbot, too. Such cases heighten concerns about safety, but don’t prove a causal explanation for the JED results.
Kayla Anderson, a psychologist and vice president of research and evaluation at the JED Foundation, explained the scope of the analysis to OPB: “We were able to look at how young people are using AI and then how that correlates separately with indicators of well-being, history of suicidal ideation and attempts, and poor mental health.”
“When young people are using AI, when they’re in emotional distress they’re more likely to report poor mental health and well-being outcomes,” Anderson said. A greater history of suicidal thoughts and attempts was also reported by those young people.
That connection doesn’t demonstrate that using chatbots caused the outcomes. “It’s unclear, based on these data, whether AI use is contributing to poor mental health and well-being, or whether students who are already in distress in certain ways are more likely to turn to AI,” Anderson said.
Maheux pointed to other research in suggesting the explanations needn’t be mutually exclusive. Young people experiencing difficulties might turn to AI for help, while their exchanges with it could make those difficulties worse instead of relieving them.
“So kids who do have worse mental health are seeking out AI to help them,” she told OPB. “But that AI use may be exacerbating those mental health challenges instead of helping.”
Dependence presents a further concern. Used as emotional training wheels, chatbots might become difficult enough to leave behind that users lose chances to build coping skills themselves. Repeated reassurance could replace practice in handling daily stress, leaving young people less able to manage on their own. Whether this is happening isn’t something the survey establishes. It does, however, leave open a meaningful distinction: helping someone develop skills differs from providing support they feel they can’t go without.
