ABSTRACT Objectives This qualitative study examined psychotherapists' perspectives regarding the use of artificial intelligence (AI) psychotherapy chatbots in adjunct with traditional therapy. Despite broad adoption of these tools by sometimes vulnerable clients, therapists are not being consulted regarding their use for mental health support. This research explores therapists' concerns and recommendations regarding the integration of AI chatbots into clinical practice. Methods Interviews were conducted with eight psychotherapists of various modalities and experience levels. Participants viewed a demonstration of an AI psychotherapy chatbot and provided feedback. Qualitative analysis was performed on the interview data, using an interpretative phenomenological framework. Results All participants felt AI psychotherapy chatbots could be useful in adjunct to traditional psychotherapy, for specific situations, demographics and issues, in particular to support cognitive behavioural therapy (CBT). No participant felt that current chatbot technology could replace the human relationship between client and therapist. Overall, six superordinate themes emerged: (1) cautious, context‐specific use of chatbots in adjunct to therapy; (2) identification of limitations and dangers, including dependency on chatbots, confusion between chatbots and humans, and increased isolation of users; (3) ethical/data privacy concerns; (4) the irreplaceable human connection; (5) the impact of chatbot use on the therapeutic alliance; (6) curious yet apprehensive attitudes toward AI. Conclusions AI psychotherapy chatbots may offer value within specific contexts, especially for CBT‐based work. However, concerns regarding ethics, data privacy, over‐reliance, loss of cognitive skills, and loss of human connection highlight the urgent need for regulation.
Aisling Mackey (Sun,) studied this question.