“Get closer!” my son said. I was standing in front of a giant piece of art that spanned half the wall. The image was completely captivating and I was reluctant to lose the perspective I had from across the room.

My kid, on the other hand, was nearly climbing into the artwork. His insistence, and the possibility that he might actually press his nose against a piece of fine art, got me moving across the room. As I walked toward him, I could see why he was so transfixed. What appeared to be a cohesive image from afar transformed into a dizzying collage of photographs, textured cloth, newspaper clips, and other carefully assembled patterns of material. 

By the time I was nose-distance away, he took my spot across the room to take in the big picture. “Woah,” he said quietly. After switching spots several more times, we eventually tore ourselves away from the piece and moved on. “I hope people get close to that one,” he whispered on our way out. 

“Far away is pretty important too,” I responded. He nodded, pleased that we had cracked the code of experiencing both.

Youth, AI, and the Value of Big Picture Data

What does this memory from the art museum have to do with kids and screens? We tend to talk about screen time from across the room, in broad brushstrokes and national averages. Headlines report the hours on screens, platform usage, and age of first device ownership. 

This is a helpful way to get a look at the big picture. It helps us see broad trends, track changes over time, and identify newly emerging patterns. Last year, when I asked parents whether they had heard of AI companions, I was usually met with blank stares. Yet a report from Common Sense Media showed that 72% of teens had used AI companions and that 52% qualified as “regular users.” Given significant and warranted safety concerns about AI companions in particular, this report helped identify an awareness gap that we needed to cross to better support young people.

In that gap, though, there are more questions than answers. Why were teens turning to AI in general? Are they really relying on AI companions? What are their motivations and needs? How are their AI habits shaped by their social relationships? How are their social relationships affected by their AI habits?  

Getting Closer to Young People’s Experiences

The Rithm Project just released a new study called Youth, AI, and the Relationships That Shape Them: Nine Portraits. This nationally representative survey of young people ages 13-24 helps us better understand the stories, motivations, and realities of teens navigating social relationships in the age of AI. The research team was committed to close-up perspectives, working closely with youth fellows to frame the survey questions, conducting follow-up interviews, and co-interpreting the findings with young people and cross-disciplinary experts.

As just one notable example, the team learned early on that the term “AI companion” carried stigma for many young people who rejected the idea of synthetic relationships. In testing, the term “AI character” was far more resonant and captured the wide range of ways young people engage with personified AI. 

They also asked young people about more than just their AI use. They asked about their social connections, sense of belonging at school, online connections, social initiative, barriers to vulnerability, loneliness, isolation, and more. In other words, they explored the rich social and emotional context of young people’s lives.

Clusters and Portraits of AI Use

What did they find? The Rithm Project team is quick to remind us that variation is the norm and that “behind every pattern of AI use is a different experience of human connection.” That said, four clusters emerged that reflected distinct uses of AI:

• Cluster 1 (28%): Uses AI Infrequently or Never
• Cluster 2 (39%): Uses AI for information and tasks
• Cluster 3 (18%): Uses AI for personal and emotional support
• Cluster 4 (15%): Talks to AI characters or personas

Digging deeper, they found important nuances within clusters. Just because young people use AI in the same way doesn’t mean they have the same motivations or needs. To pull out this detail, the team created a total of nine unique “portraits” across clusters.

For example, within Cluster 1, young people report very different reasons for not using AI. The Rithm Project team called these portraits the “conscious abstainer” (55% of Cluster 1 youth) and “AI non-participant” (45% of Cluster 1 youth). Conscious abstainers tend to distrust AI or believe it is harmful, whereas non-participants don’t know how to use AI or aren’t sure how it would help them. 

As one 18-year-old conscious abstainer reported, “I’m not retaining any knowledge whatsoever, and I don’t think [that AI is] going to help me in the long run. So, based on those two factors, I decided to decrease my activity with it.”

The research team also identified two portraits within Cluster 3: the “social processor” and the “private processor.” Social processors lean on AI for support but never or rarely turn to AI more than people when upset. Private processors, on the other hand, are likely to choose AI over people when they hit a challenge. They are also more than twice as likely as social processors to feel the urge to use AI more and more. It isn’t just AI use that drives risk but the personal and social context of its use.

Honest Realities and Complicated Storylines

There are a lot of simple storylines about young people and technology that dominate the headlines. One of the simplest is that teens plus screens equals bad outcomes. This study calls that into question. For example:

• The largest group (Cluster 2) in the study are frequent, boundaried AI users with the strongest social lives and wellbeing indicators. Within this cluster, the “intentional connectors” are thriving socially and use AI as a transactional tool rather than a replacement for social connections. 
• A much smaller cluster (Cluster 4) are assigning AI human roles and show the greatest patterns of high-risk use. 
• Meanwhile, the non-users (Cluster 1) are the most socially isolated in the study and display elevated feelings of loneliness, anxiety, and sadness. This was especially true for “AI Non-Participants.”

Does this mean that using AI fuels belonging? Not so fast. The non-user group also had the lowest socioeconomic status of all the groups. One of the key takeaways of this study is that “youth from low-income households face a compounding gap: less engagement with AI and greater social vulnerability.” When it comes to youth wellbeing, the relevant details are rarely about tech habits alone. It’s also worth noting that this is correlational data so making quick causal claims in any direction is not a good idea. 

Another Invitation to Get Curious

This study does not offer definitive conclusions about AI’s impact on teens’ social lives and wellbeing. It does offer an invitation to adults to walk closer, get curious, and be open to what new stories emerge. The Rithm Project has even created a conversation kit to help you explore the portrait cards and reflect on questions like:

• Do any of these portrait cards resonate with you or remind you of a young person you know? 
• What fits? What doesn’t fit? How might a portrait be limiting?
• What curious non-judgmental questions might you ask young people to better understand how and why they turn to AI? 
• What strengths and vulnerabilities might accompany different portraits?
• What might young people in each portrait need (from adults, systems, peers, policymakers) to navigate AI in healthy ways? 
• How does this change or shift broad generalizations about young people and AI?

Looking at the Big Picture

The rich detail that emerges when we dive into the portraits doesn’t mean we should lose sight of the bigger picture. Some of the broader insights took my breath away. Not because they related to some new AI innovation I hadn’t heard of. But because they were achingly and familiarly human. Among the Rithm Project’s main takeaways, these stood out to me:

“Almost all AI use has relational impact, but the more intimate the use, the greater the risk”

The impulse to turn to AI instead of humans is far greater among the relational support and AI character clusters than it is among task-based users. Whether or not AI complements a rich set of offline supports or replaces them is a critical question.

“Feeling genuinely seen and safe is what protects young people from high-risk AI use.”

The sheer number of connections in a young person’s life is not nearly as important as the quality of those connections. The biggest protective factor against high-risk AI use in this study was authentic relationships that make youth feel like they matter.

“There are many bright spots of young people exercising discernment with AI.”

Rather than just mindlessly adopting new technologies, this report illustrates the ways that more than half (53%) of young people across portraits are using AI relationally with intention and agency. 

“Young people are mostly navigating AI alone, without the adults in their lives.”

Over half of young people aren’t talking about AI with parents, caregivers, or teachers. When these conversations do happen, they tend to focus on academics rather than the relational use of AI. 

The Art of Staying Connected

AI innovation is moving at breakneck speed. Public concern about AI and its impact on adolescent wellbeing is keeping pace. The implications for policymakers and tech companies are clear and urgent. Advocating for safer, human-centered AI design and meaningful tech accountability matter.

And yet this study also reinforces what we have known about healthy adolescent development for a long time. Connection and conversation with caring adults matter. Belonging and safety with peers matter. Equitable access to resources and support matters. Mattering matters.

These insights aren’t new to the AI era. But we do need to translate them into relevant supports that meet the complex realities of social development in a world where Claude and Replika are always available. This will demand that we continuously stand back and take in the bigger view. And that we consistently walk toward our kids to better understand their experiences and motivations. 

As a parent of two adolescents, this push and pull feels familiar. If their lives are a work of art, sometimes I don’t even feel like I am in the same building. Other times, I get stuck in the chaotic textures right in front of me. Ultimately, teenagers don’t want our noses constantly pressed up against their lives. It’s their job to push us away, get out into the world, and explore who they are and where they belong. But they do need us to get curious and stay connected. They need us to set purposeful boundaries while resisting the pull of simple stories.

They need us to keep moving around the room for different views. It’s one way we can get a more accurate understanding of who they are and what they need from us as they grow.