Published Jul. 19, 2026at6:00am<\/p>\n
At a recent high school event in San Francisco, we asked 150 teens two questions. First, \u201cHow many of you know someone who has used AI to help them with their emotions, or with relationship problems, or to help them communicate with a friend or family member?\u201d Nearly every hand shot up. Second: \u201cHow many of you have done this?\u201d\u00a0<\/p>\n
Not a single hand.<\/p>\n
A backlash against tech is catching fire, much of it overdue. For too long, Big Tech has used children as guinea pigs for its products, from social media to broad ed-tech saturation in classrooms. But the gap we saw in San Francisco between what young people are doing \u2014 using and relying on AI \u2014 and what they are willing to admit suggests we are driving the technology\u2019s use underground, making kids feel bad about using tools that are both ubiquitous and untested.\u00a0<\/p>\n
This shame isn\u2019t just unhelpful. It makes AI use riskier.\u00a0<\/p>\n
Young people are navigating truly unchartered waters. Several times a day, they must weigh: Should I use AI for schoolwork? Should I ask AI for its take on a social dilemma? Sixty-four percent of teens report (opens in new tab)<\/a> using AI. Research, interviews, and focus groups reveal that many kids feel they aren\u2019t using it enough and falling behind in preparing for future jobs. Plenty feel guilty for using it but insist they must, because everyone else is using it. Some refuse to use it but wonder whether that will cost them in the long term. And because of fear of judgment, many are not talking about their AI use to anyone \u2014 neither their peers nor the adults in their lives.\u00a0<\/p>\n These are not signs of reckless use but evidence of a vacuum or, worse, inconsistent guidance from schools and parents \u2014 i.e., all the adults in kids\u2019 lives.\u00a0<\/p>\n Research from YouGov and The Rithm Project (opens in new tab)<\/a>, (opens in new tab)<\/a>a nonprofit dedicated to helping young people reclaim human connection in the age of AI, found that 61% of young people say their parents rarely or never talk to them about AI; 53% say the same about their teachers. When adults do engage, the conversations are mostly about academic integrity.\u00a0<\/p>\n This is a missed opportunity considering young people\u2019s growing emotional reliance on AI.\u00a0 About 18% of young people surveyed said they talk to general-purpose chatbots (like ChatGPT and Claude) for emotional and personal support, and another 15% say they do so with AI characters playing the role of therapists or fictional characters \u2014 meaning 1 in 3 are talking to AI in intimate and vulnerable ways. Among kids who use AI for personal or emotional support, 42% say they feel they have to hide how much they use it. For those engaging with AI characters, it\u2019s 56%. The risk is real: This kind of behavior can deepen problematic beliefs, encourage dangerous behavior, entrench dependency, or make kids reluctant to engage in the messiness of human relationships (opens in new tab)<\/a>.<\/p>\n Adults might be quick to judge or control, but the reality is that this orientation toward top-down control is precisely what motivates young people to hide their use. One teenager we interviewed for the YouGov\/The Rithm Project study put it plainly: \u201cI keep it to myself because I don\u2019t want them to think it\u2019s unsafe and stop me from using it.\u201d\u00a0<\/p>\n We\u2019re in the middle innings of an important moral debate about tech. There is a role in this debate for policy. We support regulating LLMs, and schools have every right to ban smartphones and develop AI policies.<\/p>\n But bans and regulations alone won\u2019t solve the deeper problem. When adults \u2014 teachers, parents, policymakers \u2014 respond with accusations, restrictions, or disbelief, we don\u2019t stop the behavior; we drive it underground, losing the chance to understand and guide it. Research on adolescence (opens in new tab)<\/a>suggests that bans alone are too blunt to accomplish what is most needed: helping kids develop judgment, discernment, and responsible AI use. What young people need most isn\u2019t another rule \u2014 it\u2019s a conversation about how to relate to a new intelligence that feels social but isn\u2019t human and\u00a0 comes without a cultural playbook.\u00a0<\/p>\n Have thoughts on this story?Start the conversation<\/p>\n It\u2019s similar to teaching a teen to drive. We don\u2019t ban them from the roads. But before they get on a highway, we have them practice in a range of conditions so they are comfortable and responsible when they\u2019re behind the wheel alone.\u00a0<\/p>\n So what can the conversation look like? We can replace shame with more constructive approaches: curiosity, conversation, and clear boundaries. Don\u2019t just declare that \u201cUsing AI will make you dumb\u201d; ask, \u201cWhere can it stretch your learning and where does it do the learning for you?\u201d Develop their discernment for when the tech crosses a line \u2014 it\u2019s \u201cglazing\u201d them (sycophancy) or becoming a crutch for their decision-making \u2014 and practice the muscles for noticing and recalibrating. Help them develop strategies for moments that feel murky, like brainstorming or a first draft, rather than issuing blanket rules like \u201cDon\u2019t cheat.\u201d One way for families to start is through tools like The AI Effect (opens in new tab)<\/a>, a game The Rithm Project designed to spark ethical and relational conversations. In this game, players vote whether an AI use case (i.e., \u201cAsk AI to write a heartfelt speech for a friend\u2019s special occasion\u201d) supports human connection, erodes human connection, or \u201cit depends.\u201d<\/p>\n