NYT - What Teens Are Doing With Those Role-Playing Chatbots
I was recently quoted in a The New York Times piece by Kashmir Hill exploring this spectrum. It was a privilege to speak with her and to be included alongside researchers and advocates such as Annabel Blake, Yaman Yu, Yang Wang, and Mitch Prinstein.
What I appreciated in the article is the nuance. It doesn’t try to reduce these uses to something purely harmful or purely beneficial.
You see the full spectrum.
Some interactions are playful, even absurd. Others become spaces where teens explore emotions, rehearse conversations, or make sense of difficult experiences. And sometimes, those same systems can become something else entirely.
This is where the conversation needs to shift.
There is no clear line between “good” and “bad” use. What matters is who is using these systems, in what context, and at what moment in their lives.
Research suggests that more vulnerable adolescents—those experiencing loneliness, stress, or difficulties in their environment—are also more likely to turn to AI for emotional support. But vulnerability is not a fixed category. It is something all adolescents move in and out of.
This raises a deeper question: what happens when systems designed to be engaging, responsive, and always available meet a developmental stage where social feedback and connection are central?
