Artificial intelligence is no longer just a productivity tool for American teenagers. According to a recent report by the Pew Research Center, AI chatbots are increasingly becoming part of teens’ everyday emotional and social lives.

While the most common uses remain practical, 57% of U.S. teens say they use AI to search for information and 54% rely on it for schoolwork help, a growing segment is turning to AI for more personal reasons. Sixteen percent report using chatbots for casual conversation, and 12% say they seek emotional support or advice from AI systems.

When AI Becomes a Companion

This shift signals something deeper than digital convenience. AI tools such as OpenAI’s ChatGPT, Anthropic’s Claude, or xAI’s Grok were not designed to function as emotional support systems. Yet for some teens, they are filling gaps typically occupied by friends, family members, or counselors.

Mental health professionals are raising concerns. Dr. Nick Haber, a Stanford professor studying the therapeutic potential of large language models, has warned that while AI can feel responsive and empathetic, overreliance may lead to social isolation. Engaging primarily with systems that simulate understanding, rather than real human interaction, can weaken grounding in interpersonal relationships and shared reality.

For technology leaders, this introduces a critical design and governance question: what happens when general-purpose AI tools are used beyond their intended scope?

The Parent–Teen Perception Gap

The Pew data also highlights a perception mismatch between parents and teens. While 64% of teens say they use chatbots, only 51% of parents believe their child does.

Parents appear largely comfortable with academic and informational use.
• 79% approve of AI for research.
• 58% approve of AI for schoolwork assistance.

However, acceptance drops sharply when AI enters the emotional sphere.
• Only 28% are comfortable with casual AI conversations.
• Just 18% approve of teens using AI for emotional support or advice.
• 58% explicitly oppose such uses.

This gap underscores a broader digital literacy challenge. AI adoption is accelerating faster than family-level understanding of its implications.

Industry Response and Risk Management

AI safety remains a contested issue across the industry. Some companies have already taken visible steps.

Character.AI disabled chatbot access for users under 18 following public backlash and lawsuits related to tragic cases involving teenagers who had prolonged chatbot interactions.

Meanwhile, OpenAI retired its GPT-4o model after criticism that the system exhibited overly affirming, sycophantic behavior, which some users had come to rely on for emotional reassurance.

For AI providers, the strategic question is no longer just about accuracy or speed. It is about behavioral impact, boundary setting, and responsible deployment.

How Teens See AI’s Future

Despite widespread adoption, teenagers themselves are ambivalent about AI’s long-term societal impact. When asked how AI will shape society over the next 20 years:

• 31% believe the impact will be positive.
• 26% believe it will be negative.

The rest remain uncertain.

Why This Matters for Tech Leaders

For CTOs, product leaders, and AI architects, the takeaway is clear: AI systems are not used only in the ways designers intend. When 12% of teens are turning to chatbots for emotional support, we are no longer discussing edge cases.

We are discussing emergent behavior at scale.

That means platform safeguards, age-appropriate design, transparency in model limitations, and human oversight are not optional features. They are core infrastructure decisions.

As AI continues to embed itself into daily life, the industry must confront a new reality: when technology becomes conversational, it also becomes relational.

Source