Psychology researcher talks about teens and generative AI

In graduate school, Anne Maheux was interested in broadly studying the social and cultural experiences of children and adolescents. It wasn’t until late in her graduate studies that she realized she’d become a person who studies the digital world.

It was a lightbulb moment for Maheux, assistant professor of psychology and neuroscience in the UNC College of Arts and Sciences and Winston Family Distinguished Fellow at the Winston Center on Technology and Brain Development.

She recognized that the sociocultural and digital worlds are blended for kids today, and there is no way for researchers to fully understand the social world of adolescents without considering technology, including artificial intelligence.

Modern digital media have been adopted at a speed that far outpaces the research into their effects. Maheux and her team are launching a new longitudinal study to better understand the complex ways in which digital risks impact different kids in different ways over time.

“The adoption pace for generative AI has been shockingly fast, faster than that of social media, personal computers or the internet itself. And adults are not the only ones using AI,” Maheux said. “Our pilot data has indicated that 20% of 10- to 12-year-olds and 42% of 13- to 14-year-olds are using generative AI. The actual numbers are likely even higher, as data only measures use on personal mobile devices and doesn’t account for time spent on computers or shared devices.”

Youth mental health concerns are on the rise, and numerous studies have indicated an association between digital media use and body image issues, eating disorders, depression, anxiety and other mental health concerns for adolescents.

“We know that kids have different levels of vulnerabilities. From the research I’ve seen and led, I believe that adolescents in the 10- to 14-year age range are the most vulnerable to threats in the digital world,” Maheux said. “This group is independent enough to access content on the internet, often even as their parents actively try to limit access, but most don’t yet have the coping skills to process everything they encounter online.”

The different motivations that an adolescent has for using these tools is also a very important factor that leads to different outcomes.

“For instance, we suspect that young people who are feeling lonely at school and turn to an AI chatbot for friendship or romance are more likely to have poor mental health outcomes associated with their use of the technology compared to students who use AI chatbots as a means of learning new skills or exploring topics of interest,” Maheux said.

What can parents do to help their children?

“I’m not a parent and I don’t claim to know the best way to approach technology for all families, but I also don’t think parents have to wait until the research is perfect to decide they should do things differently for their children,” Maheux said.

But she does have some basic advice.

“Before parents hand a device to their child, I think it’s helpful to talk to them, in specifics, about what will be done on the device. And it’s important for parents to ask their kids what they are seeing online so they can provide support if their child is seeing anything they don’t understand,” Maheux said. “One of the worst things that can happen to kids today is having a challenging experience online and not feeling like they can talk about it with a trusted adult.”