Artificial intelligence (AI) chatbots have become widely accessible and are increasingly used by the general public for emotional support, decision-making, and health-related inquiries. Although these tools hold considerable potential for enhancing mental health care in structured settings, their unsupervised use by individuals with psychiatric conditions could present underexplored risks. Drawing on theoretical and emerging empirical literature, this article suggests how AI chatbot interactions might unintentionally reinforce some transdiagnostic psychiatric patterns, including delay in help-seeking, excessive reassurance-seeking, social withdrawal, distorted cognitions and perceptions, and reduced autonomy through overreliance. This article calls for further empirical research and increased clinical and ethical scrutiny of public chatbot use among vulnerable individuals and suggests potential safeguards for both developers and clinicians to ensure that AI integration into society remains both beneficial and safe. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Javad Abbasi Jondani (Mon,) studied this question.