JMIR Publications

ABSTRACT

Work in embodied artificial intelligence (AI) has increasing clinical relevance for therapeutic applications in mental health services. With innovations ranging from ‘virtual psychotherapists’ to social robots in dementia care and autism disorder, to robots for sexual disorders, artificially intelligent virtual and robotic agents are increasingly taking on higher-level, therapeutic interventions that used to be offered exclusively by highly trained, skilled health professionals. Embodied AI applications in mental health care carry hopes of improving quality of care and controlling expenditure, as well as reaching underserved populations in need of mental health services, and improving life opportunities for vulnerable groups. However, there is a persistent gap between current, rapid developments in AI mental health, and the successful adoption of these tools into clinical environments by health professionals and patients. In addition, interventions are often designed without any explicit ethical considerations. At present, the quality of research on embodied AI in psychiatry, psychology, and psychotherapy is varied, and there is a marked need for more robust studies, including RCTs on the benefits and potential harms of current and future applications. In light of the demonstrated potential, we provide an analysis of ethical and social implications of currently emerging embodied AI applications in mental health. We argue that embodied AI is a promising approach across the field of mental health, however, further work is needed to address the broader ethical and societal concerns of these technologies in order to negotiate best research and medical practices in innovative mental healthcare. We close by indicating areas of future research and developing recommendations for high-priority areas in need of concrete ethical guidance.