This article is based on my latest Youtube Video: AI, Anger & Teens: What Happens When We Practice Aggression on Chatbots? by Dr. Élena Bagourdi
For years, concerns about technology and adolescence have centered on social media. Researchers and parents have worried about cyberbullying, online comparison, and the emotional effects of constant digital connection. But as conversational AI becomes increasingly integrated into everyday life, a different psychological question is beginning to emerge: What happens when teenagers practice anger in relationships that feel human—but aren't?
Recent reports suggest that a significant portion of interactions with conversational AI involve verbal aggression, insults, or even violent role-play. While these exchanges may seem harmless because no person is directly harmed, they raise important questions about emotional development. Aggression itself is not the problem. In psychology, aggression is not synonymous with violence or cruelty. Healthy aggression helps young people develop independence, establish boundaries, advocate for themselves, and explore their identity. A toddler saying "no," a teenager disagreeing with parents, or a young person standing up to peer pressure are all examples of adaptive aggression. The challenge is learning how to regulate those impulses.
Historically, adolescents developed emotional maturity through real relationships. When they hurt someone's feelings, they witnessed the consequences. They experienced disappointment, conflict, repair, forgiveness, and accountability. These experiences gradually taught empathy, frustration tolerance, and the understanding that other people have their own thoughts, feelings, and boundaries. Conversational AI changes that learning environment. Unlike human relationships, AI systems often remain available regardless of how they are treated. They can absorb frustration, be insulted without protest, and instantly restart conversations without lingering emotional consequences.
While AI can be a helpful educational and creative tool, repeated interactions without genuine relational feedback may reduce opportunities to practice one of life's most important developmental skills: repairing relationships after conflict. This does not mean that using AI causes emotional problems. Nor does it mean AI is inherently harmful. Instead, it highlights the importance of understanding how different kinds of relationships shape development.
Human relationships ask us to tolerate discomfort, negotiate differences, and recognize the emotional impact of our behavior. These experiences help cultivate empathy, accountability, and emotional regulation. For parents, the goal is not punishment or shame.
If you discover your child interacting aggressively with AI, approach the conversation with curiosity rather than criticism. Ask what they enjoy about the interaction. Explore what emotions they are expressing and whether those feelings also arise in relationships at school, home, or with friends. It is also worth reviewing AI settings. Many experts recommend limiting personalization and memory features for children and adolescents, reducing the likelihood that an AI system becomes an ongoing pseudo-relationship during a critical stage of social development.
Most importantly, if a young person is struggling with anger, loneliness, anxiety, or social difficulties, those concerns deserve attention beyond the AI itself. Sometimes technology is not the problem—it simply becomes the place where deeper emotional struggles are expressed.
As AI continues to evolve, one of the defining psychological questions of the coming decade may not be whether young people use artificial intelligence, but how those interactions influence the way they learn about anger, intimacy, empathy, and human connection. Technology will undoubtedly remain part of adolescence. Our challenge is ensuring that it complements, rather than replaces, the real relationships that teach young people how to navigate the complexities of being human.
Explore more of my work: