Abstract
Individuals are increasingly using artificial intelligence chatbots, such as ChatGPT, to seek conversational support for their personal mental health difficulties. Heightened concerns about mental health stigma may make anonymous, on-demand chatbot interactions more appealing for some than traditional face-to-face support. This study examined if using ChatGPT-4 for personal mental health difficulties is associated with two distinct forms of stigma, anticipated stigma and self-stigma. Our main aim was to investigate if the perceived effectiveness of ChatGPT use for mental health issues mediates the relationship between ChatGPT usage and anticipated stigma and self-stigma. The sample comprised 73 participants, mostly undergraduate psychology students. Participants completed online self-report measures to assess ChatGPT usage for mental health purposes, perceived effectiveness of ChatGPT for mental health issues, and anticipated stigma and self-stigma. Perceived effectiveness of ChatGPT was significantly and positively correlated with ChatGPT usage, and significantly negatively correlated with reduced anticipated stigma. Cross-sectional analyses found that perceived effectiveness significantly mediated the relationship between ChatGPT use and anticipated stigma, but not for self-stigma. The results indicate that ChatGPT use, when perceived as effective, is associated with a reduction in anticipated stigma concerning mental health issues. More research is now needed in this emerging area to inform best practice on the use of AI aids for mental health issues.