Therapists account for about 8.2% of patient outcomes – that is, some therapists are reliably more...
Therapists account for about 8.2% of patient outcomes – that is, some therapists are reliably more...
The therapeutic alliance is a well-known and reliable predictor of patient outcomes in...
The public health restrictions related to the COVID-19 pandemic are over, and it is time now to...
There are multiple theories of the role of interpersonal problems in depression. One theory holds...
PPRNet Clinical summary: The Most Vulnerable are Prone to Use AI Therapists
Digital mental health applications include artificial intelligence (AI) tools such as chatbots and avatars that mimic human therapists. Human-based psychotherapy is grounded in the healing aspects of relationships, such as empathy, therapist presence, and a collaborative therapeutic alliance. However, some people may be drawn to AI tools (chatbots and avatar therapists) for reasons that reflect their relational vulnerabilities. For example, those who struggle with interpersonal relationships may avoid closeness or may not know what it is like to be held emotionally and feel understood. Such individuals may turn to AI-based interventions to create a pseudo-connection that helps them to circumvent the relational challenges they face. In this study, Bekes and Aafjes-van Doorn examined how a person’s psychological profile, including level of mental health symptoms, epistemic trust (openness to opportunities for social learning in the context of relationships), and attachment security, predicts acceptance of AI-based interventions versus human-delivered therapy. The authors conducted a survey of 1,612 patients and clinicians. They conducted a cluster analysis of the psychological data to identify three groups of participants: (1) Attachment avoidant-not trusting, (2) Securely attached-trusting-healthy, and (3) Anxiously attached-credulous-symptomatic. The Attachment avoidant-not trusting and the Anxiously attached-credulous-symptomatic groups had significantly higher acceptance of AI-based interventions. In contrast, the Securely attached-trusting-healthy group showed the lowest acceptance of AI-based interventions. Also, those who had previous experience of psychotherapy with a human had the lowest acceptance of AI-based interventions.
Practice Implications
The authors offered an optimistic interpretation of the findings, suggesting that those with relationship problems (avoidantly or anxiously attached, and epistemically mistrusting or credulous) might use AI-based interventions to access psychotherapy. However, one could also argue that AI-based therapies may reinforce problematic relational patterns. Those with attachment avoidance and epistemic mistrust may continue to avoid human relationships through AI-based therapies. Those with attachment anxiety and epistemic credulity may be comforted by the 24/7 support of a chatbot, leaving little room for developing healthier reliance on their own capacities.
By Dr. Giorgio Tasca
The director of the Psychotherapy Practice Research Network (PPRNet) is Dr. Giorgio Tasca. Dr. Tasca is an Associate Professor with the School of Psychology, in the Faculty of Social Sciences at the University of Ottawa. His research is centered around psychotherapy process, mechanisms of change, and outcomes, as well as eating disorders.Also Read