Perceptions of Therapists Identified as
AI
Generated or Human
Jasmine R. Blake, Jeffrey S. Berman, Erin M. Solomon ABSTRACT
Background
As artificial intelligence (AI) continues to advance, its integration into mental healthcare offers both promise and uncertainty. Digital therapists have shown effectiveness in reducing symptoms of anxiety and depression, yet scepticism remains about AI's ability to provide human‐like empathy and authenticity.
Aims
This study aims to assess whether participant perceptions differ based on whether a therapist is described as AI generated or human.
Materials and Methods
Participants ( N = 300) listened to two therapist audio recordings identical in content and both from a human therapist. However, in one of these recordings the therapist was identified as AI generated; in the other recording the therapist was identified as human. After listening to each recording, participants evaluated therapist empathy, perceived credibility of treatment and expectation for treatment success.
Results
Analysis revealed that participants judged therapists labelled as human more positively across all measures. Although the results did not vary reliably as a function of therapist sex, prior therapy experiences moderated perceptions when considered alongside therapist sex.
Discussion and Conclusion
These findings suggest that the belief that the therapist is AI results in less favourable perceptions even when actual therapy content is identical, revealing ongoing challenges for the acceptance of artificial intelligence within the field of psychotherapy.