The Myriad Complex Ways Young People Use AI

Today’s young people are the first generation to grow up with artificial intelligence systems during a formative period of building identity and connection, and many turn to AI for emotional support and social interaction. But a new report from The Rithm Project finds a more nuanced picture of how they engage with the technology.

Last fall, the organization surveyed nearly 2,400 people ages 13 to 24—including more than 500 young adults ages 18 to 24—on how their AI use intersects with their social lives, emotional well-being and relationships. The data, collected in partnership with YouGov, was weighted to be nationally representative across age, gender, race and geography in the United States.

Respondents fell into four groups: 28 percent said they use AI infrequently or not at all, 39 percent use it for information and tasks, 18 percent use it for personal or relational support, and 15 percent said they interact with AI characters—the most intimate form of use identified in the study, which defined it as ongoing engagement with “humanlike” systems.

Alison Lee, chief research and development officer at The Rithm Project, a nonprofit dedicated to fostering meaningful human connection in the age of AI, said that people may be overestimating the number of young people engaging in high-risk AI companionship while underestimating how AI is shaping their relationships and sense of connection.

“We need to be more precise about the spectrum of how young people are engaging with AI—and, more importantly, ask why,” she added. “What we’re finding is that their human relationships are really shaping how they enter into AI use.”

Laura Erickson-Schroth, chief medical officer at The Jed Foundation (JED)—which contributed to the research with The Rithm Project—said the way young people approach AI mirrors how earlier generations encountered social media, when they pioneered its use and their parents tagged along.

“There’s a generation of adults encountering AI tools at the same time as their children—similar to what happened with social media,” Erickson-Schroth said. “That makes it harder for adults to give guidance, set boundaries and have meaningful conversations, because they don’t fully understand the context.”

Patterns of AI use: The four groups were further divided into nine distinct portraits of young people, the report said.

“What we realized, even within this first pass of AI usage, is that similar behaviors can mask very different underlying motivations and relational contexts,” Lee said. “That’s where the portraits emerged. We tried to go a layer deeper and ask what relational environments and social supports might be shaping how young people arrive at the same kinds of AI use—but for very different reasons.”

Among those who use AI infrequently or not at all, two profiles emerged. “The Conscious Abstainer,” representing 55 percent of this group, tends to distrust AI or view it as harmful, while “The AI Non-Participant,” making up the remainder, demonstrates limited understanding of how to use AI or uncertainty about its value.

Researchers also identified two additional portraits among those who use AI for information and tasks. “The Intentional Connector,” representing 53 percent of the group, reports strong social networks and rarely feels lonely, often prioritizing in-person interactions. By contrast, “The Brave Face,” at 47 percent, maintains moderate social connections but still experiences loneliness and isolation.

For those using AI for personal and relational support, the report again identified two groups. “The Social Processor,” representing 53 percent, typically uses AI as a tool for reflection or rehearsal rather than relying on it as a primary source of support. “The Private Processor,” at 47 percent, is more likely to turn to AI over other people when dealing with difficult emotions, suggesting the technology may be filling a gap in their support systems.

Among the most intensive users—those interacting with AI characters—the report identified three portraits, with some respondents fitting more than one. “The Bestie,” representing 60 percent, engages with AI in deeply personal ways, treating it as a friend, family member or romantic partner. “The Gamer,” at 54 percent, uses AI for creative role play and interaction with fictional characters, often demonstrating high or compulsive use. “The Expert Seeker,” representing 47 percent, turns to AI for guidance in roles like therapist or coach, valuing its accessibility but also showing signs of reliance.

Other research reinforces this broader pattern. A study from Surgo Health found that about 12 percent of young people reporting mental health challenges said they turned to generative AI to talk through their feelings. While short-term emotional relief was common, the study found it did not consistently translate into positive long-term outcomes. When AI was used alongside broader support systems, benefits were more likely. But when it functioned as a substitute for human connection, outcomes were more often neutral or negative.

“The biggest predictors of high-risk AI use are when young people feel they can’t be real with the people around them, feel like a burden when they share their problems and feel like they have no one to turn to when they need help,” Lee, of The Rithm Project, said. “Conversely, the strongest protective factors are having people who make them feel they matter, and relationships where they can be their unfiltered selves.”

“That has major implications for the institutions that hold human relationships, like colleges,” she added. “How do we build environments where young people feel seen, feel safe being themselves and feel like they matter—so they’re not afraid to ask for help?”

Erickson-Schroth agreed, noting that “young people are really smart—once you start talking about these ideas, they engage quickly, ask questions and want to understand how these systems are built.”

“Adults don’t always see the ways that young people are actively making choices about their engagement with technology,” she added. “Many young people are really discerning about how they use AI … and I think that’s what’s missing from the conversation.”

The bigger picture: The nine portraits “represent a snapshot in time rather than fixed categories,” the report notes, pointing to larger trends in how young people engage with AI and suggesting new portraits could emerge as the technology evolves.

The findings also carry implications for the adults and institutions shaping young people’s AI use.

For mental health professionals, that means moving beyond simply asking whether they use AI to understand the role it plays in their emotional lives. For policymakers, the report calls for safer, less “anthropomorphic” AI systems and regulations that reward platforms for strengthening real-world connection, skill building and help seeking.

For educators, the report urges shifting conversations beyond academics and cheating to acknowledge the personal and relational ways AI is already shaping students’ lives. And for parents, mentors and family members, the report underscores the importance of approaching young people with curiosity and without judgment—suggesting that open, nonjudgmental conversation may be one of the most important interventions available.

“Where I see real opportunity is that there are a lot of adults who are experts in this area, who know a lot about this emerging technology and can help to build digital literacy programs for schools,” Erickson-Schroth said. “I do think it’s really important that schools have digital literacy programs that include AI and how young people can navigate AI because we want to make sure that they feel like they’re in the driver’s seat and can make choices with all the information that they need.”

Get more content like this directly to your inbox. Subscribe here.