0
0
February 12, 2025—A new study published in PLOS Mental Health suggests that ChatGPT, a generative AI model, has the potential to enhance psychotherapeutic processes. The research, conducted by H. Dorian Hatch of The Ohio State University and co-founder of Hatch Data and Mental Health, found that AI-generated responses were often rated higher than those written by professional psychotherapists.
The study explored whether people could differentiate between responses written by human therapists and those generated by ChatGPT. With over 800 participants assessing 18 couple’s therapy vignettes, the results showed that while some differences in language patterns were noted, individuals rarely identified the source of the responses correctly. This finding aligns with Alan Turing’s long-standing prediction that humans might struggle to distinguish between human and AI-generated text.
Moreover, ChatGPT’s responses were consistently rated higher in core psychotherapy guiding principles. Further analysis revealed that AI-generated responses tended to be longer and contained more nouns and adjectives than those written by therapists. These linguistic elements may have contributed to the AI’s ability to provide extensive contextualization, which could explain the higher ratings from participants. Researchers suggest that this feature of generative AI may enhance the common factors of therapy, fundamental components that drive successful therapeutic outcomes.
The findings indicate that AI tools like ChatGPT could be leveraged to develop new methods for testing and creating psychotherapeutic interventions. Given the growing evidence supporting AI’s utility in mental health settings, the study’s authors urge mental health professionals to expand their technical literacy to ensure AI models are responsibly trained and supervised. This, they argue, could improve both the quality and accessibility of mental health care.
Reflecting on the broader implications, the researchers highlighted the historical debate on whether AI could serve as a therapist since the development of the early chatbot ELIZA nearly sixty years ago. They acknowledge that critical ethical and feasibility questions remain but emphasize that their findings suggest AI could indeed play a role in psychotherapy.
“We hope our work galvanizes both the public and mental health practitioners to ask important questions about the ethics, feasibility, and utility of integrating AI in mental health treatment before the AI train leaves the station,” the authors stated.
Disclaimer: While the study suggests potential benefits of AI in psychotherapeutic settings, it does not advocate replacing human therapists with AI. AI should be viewed as a supplementary tool rather than a substitute for professional mental health care. Ethical considerations, data privacy, and human oversight remain crucial factors in AI’s integration into therapy. Readers should consult licensed professionals for mental health concerns.
