A 2025 survey of over 1,000 US teens found 72% had tried AI companions and 52% used them regularly, but the detail that unsettled researchers was this: a third had turned to a bot, not a person, for a serious conversation

A 2025 survey of more than 1,000 US teens indicates that 72% have utilized AI companions, with 52% reporting regular usage. Researchers have expressed concern that one-third of these teens have chosen to discuss serious matters with a bot rather than a human. This trend now extends to clinical settings, where patients are increasingly bringing AI tools into their therapy sessions.

Coverage from Employee Benefit News, Solutions Review, the Pulitzer Center, Neuroscience News, the American Psychological Association (APA), and Silicon Canals highlights a growing focus on the mental health risks associated with AI. Reports emphasize the shift from simple advice-seeking to the formation of AI-based relational bonds, with industry outlets analyzing the limits of technology in addressing broader mental health crises. Future developments remain dependent on further investigation into the impacts of chatbot interaction on mental well-being.

Observers are monitoring how professional organizations, such as the APA, will address the integration of AI into traditional therapeutic practices.