A Comparison of Responses from Human Therapists and Large Language Model–Based Chatbots to Assess Therapeutic Communication: Mixed Methods Study

Summary & key facts

The American Psychological Association wrote a short policy saying psychologists should help shape how artificial intelligence (AI) is made and used. They explain that AI is already part of school, work, and healthcare, and that psychology has special knowledge about how people think and feel. The policy lays out three main roles for psychology: influence how AI affects society, help use AI to improve health and access to care, and make sure AI follows ethical and privacy rules. It also lists practical priorities, like reducing bias, doing more research, teaching people how to use AI wisely, and advising policy makers and tech developers so AI helps people and avoids harm.

Key facts:

• AI is already being used a lot in everyday places like healthcare, school, and work, so its effects on people matter now.
• The American Psychological Association says psychology should play three big roles: shape AI’s effects on society, help AI promote health and wellbeing, and address ethics and privacy issues around AI.
• The policy says psychology’s knowledge about how people think and behave is needed to reduce harm and bias in AI systems.
• The APA lists specific priorities: center ethics and human rights, promote equity and reduce bias, include diverse perspectives in AI design, and keep researching how AI affects people.
• The policy says AI could help increase access to care and early intervention, but those benefits will only happen if psychology is included in research and development.
• The APA calls for teaching the public how to tell AI from real people and how to use AI tools safely and responsibly.
• The document does not claim AI is already safe or ready to replace human care; it calls for ongoing study, critique, and careful evaluation before AI is widely used in sensitive areas.

Abstract

Our study demonstrates the unsuitability of general-purpose chatbots to safely engage in mental health conversations, particularly in crisis situations. While chatbots display elements of good therapy, such as validation and reassurance, overuse of directive advice without sufficient inquiry and use of generic interventions make them unsuitable as therapeutic agents. Careful research and evaluation will be necessary to determine the impact of chatbot interactions and to identify the most appropriate use cases related to mental health.

Topics

AI in Service Interactions Digital Mental Health Interventions Mental Health via Writing

Categories

Applied Psychology Psychology Social Sciences

Tags

Applied psychology Artificial intelligence Chatbot Computer science Human–computer interaction Intervention (counseling) Medical education Medicine Mental health Psychiatry Psychological intervention Psychology Psychotherapist Qualitative research Social science Sociology Thematic analysis Think aloud protocol Usability

Conditions & symptoms

Anxiety Depression