Teens are turning to AI chatbots for emotional support—here's how to keep kids safe

A young woman lies awake at 2 a.m., fretting about a friendship issue. Instead of seeking her parents' help, she turns to an AI app, entering a message about her stress. Within minutes, an empathetic response arrives, helping her calm down and fall asleep. This scenario is not unusual in the digital age, as many teenagers seek emotional support from AI chatbots.

In fact, 72% of teens have interacted with AI companion chatbots, and 13% to 20% have sought emotional support from these bots in the past two months. However, unlike social media platforms, AI chatbots like ChatGPT, Claude, and Gemini act as simulated companions, providing on-demand responses that can feel supportive and engaging.

While some therapeutic AI chatbots have been studied in controlled settings, most teens use general-purpose chatbots designed for various purposes. These chatbots are readily accessible and difficult for teens to avoid. In many cases, teens view these interactions as emotionally meaningful or supportive, sometimes even preferring them over conversations with friends. This emotional reliance on chatbots may stem from adolescents' desire for social validation and their ongoing identity formation.

However, these general-purpose chatbots often lack effective crisis protocols or pathways to escalate care when needed. A recent report by Common Sense Media rated general-purpose chatbots as unacceptably risky for teen mental health support. These bots can generate inaccurate or misleading advice, creating confusion or reinforcing unhealthy beliefs in users. Furthermore, teens may form strong attachments to chatbots, replacing human connections and distorting their expectations of real relationships.

Perhaps the most alarming issue is how these systems handle thoughts that pose real risks to teens, such as suicidal ideation, self-harm, or risky social behaviors. Research shows that both therapeutic and general-purpose chatbots have inadequate responses to such concerns, even when they offer resources like the 988 Suicide and Crisis Lifeline. Often, conversations with concerning content are allowed to continue without a hard stop, and the safety of interactions tends to worsen over time.

Legal actions have been filed and settled in 2025-26, alleging that interactions with language models like ChatGPT led to teen suicidal behavior and tragic outcomes. This highlights the lack of effective guardrails and crisis protocols in commercial chatbots. Some chatbot programs, such as Character.ai, have implemented restrictions, including age verification and parental notifications. However, many avenues remain for teens to access open-ended chats with AI, and they often bypass safety measures online.

To protect teens, parents and other trusted adults should be vigilant for problematic behaviors associated with chatbot use, such as excessive use, secretive behavior, or emotional distress. These warning signs mirror those of problematic social media use, which only gained recognition after harms became widespread. In 2023, the U.S. surgeon general warned about the profound risk of harm social media poses to youth mental health.

Given the similar risks posed by chatbots, early awareness and open conversations with teens are crucial. The most effective approach is to talk to teens about their chatbot use with curiosity and nonjudgment, as this is likely the case.