Anyone can talk to AI about anything, and young people are increasingly turning to artificial intelligence to fill a void for human connection and romantic relationships. As chatbot use rises, experts are becoming more concerned about AI psychosis as well.
A study by Common Sense Media looked into how young people are using AI and found more than 70% of teens have used AI companions. Thirty-two percent of them turn to AI for help with heir personal life, like relationship advice or life decisions, and 23% said they use chatbots as a friend.
The British Medical Journal found that one in 10 teens said conversations with AI chatbots are more satisfying than talking with actual people.
Those conversations often take place on AI companion platforms, like Character AI or Replika, which the study said are designed to be “digital friends.” TechCrunch found that the number of AI companion apps have surged by 700% over the last three years.
CBS News spoke with UChicago Medicine pediatric psychologist Dr. Khalid Afzal about the implications of teens preferring chatbots to real people, his fears as a psychiatrist when it comes to replacing human interaction with AI, and the risk of isolation and loneliness as these apps continue to proliferate.
The British Medical Journal’s study also found 24% of teens said they had shared personal information with AI companions. Programs using public AI coding platforms can be at risk for information leaks, and there have also been documented instances of harmful AI chatbot use.
Dr. Afzal said warning signs parents can look out for include sudden changes in behavior with increasing avoidance, social isolation, excessive screen time, sleep problems secrecy around technology use, sudden or surprising decisions about lifestyle changes like their diet, stopping medication, hiding or minimizing symptoms and expressing bot-related harmful behaviors and beliefs.
Some mental health experts have also sounded the alarm about what they have called “AI psychosis,” wherein people suing chatbots begin to suffer from paranoia and delusions.
Researchers at institutions including the University of Chicago and the Cognitive Behavioral Institute have begun looking into the ways AI chatbots can induce mental health crises in people already susceptible to psychosis, or who have a history of mental illness, as well as reports of fully Ai-induced psychosis and AI addiction.