Alignment of Large Language Model Responses With Human Therapists in Motivational Interviewing
In this cross-sectional study of 154 MI sessions, prompted LLMs showed general alignment with therapist responses in MI-oriented conversations, as judged by automated similarity metrics. However, limitations in long-range coherence, stylistic alignment, and the use of indirect proxies for therapeutic quality highlight the need for improved prompt design, MI-specific evaluation methods, and clinical validation before integration into mental health care.
Authors
- Richard J. Zeifman (ORCID: https://orcid.org/0000-0003-3478-4483)
- Candice M. Monson (ORCID: https://orcid.org/0000-0001-6179-0788)
- Alice Rueda (ORCID: https://orcid.org/0000-0002-9977-9653)
- Bazen Gashaw Teferra (ORCID: https://orcid.org/0000-0001-5325-9639)
- S. Huang
- Karisa Parkington
- Nabil Johny
- Sri Krishnan
- Argyrios Perivolaris
- Venkat Bhat
- Huda Al-Shamali
- Divya Sharma
Institutions
- The New School (US)
- St. Michael's Hospital (CA)
- University of Waterloo (CA)
- University of Toronto (CA)
- York University (CA)
- St Michael’s Hospital (IE)
- Imperial College London (GB)
- Toronto Metropolitan University (CA)
- New York University (US)
- McMaster University (CA)
Publication Details
- Journal
- JAMA Network Open
- Published
- 2026-03-23
- DOI
- https://doi.org/10.1001/jamanetworkopen.2026.2750
- Primary Topic
- Digital Mental Health Interventions
- Type
- article
- Field-Weighted Citation Impact
- 0.00