Artificial intelligence robots for mental health applications: a scoping review

The contradiction between the surging demand for mental health services and the shortage of professional resources is becoming increasingly prominent. Artificial intelligence robots are a promising tool for digital mental health interventions, offering unique application potential in this field.

The purpose of this review is to systematically analyze the characteristics and technical pathways of AI-robot in mental health, to help developers make sound decisions in technology selection, and to provide a basis for mental health providers and researchers to select suitable robots.

A scoping review was conducted using the JBI method and reported according to the PRISMA extension for scoping reviews (PRISMA-ScR). We searched PubMed, Embase, PsycINFO, Scopus, IEEE Xplore, ACM Digital Library, and Cochrane Central Register for Controlled Trials database to collect research literature on AI-robot applications in mental health published between January 1, 2017, and May 5, 2025, and used a narrative synthesis approach to integrate the extracted evidence.

A total of 34 studies were included. AI-robot were used primarily for diagnosis (n=19), monitoring (n=5), treatment and intervention (n=27), risk prediction and identification (n=8), and counseling (n=13). The research focuses on anxiety disorders, depressive disorders, autism spectrum disorders, and dementia groups. In diagnostic applications, machine learning algorithms are the core technology, combined with multimodal data processing, that can enhance diagnostic efficacy. In therapeutic applications, natural language processing technology serves as the core technology. Combined with generative AI, it can significantly enhance the potential of personalized intervention and effectively alleviate mental health issues. The results of the AI performance metrics show that the relevant models perform well overall, with accuracy being the most frequently reported evaluation metric.

Although the application of AI-robot in mental health services is promising, significant challenges remain in terms of clinical efficacy verification, technical effectiveness, privacy protection, and ethical compliance. Future studies need to follow the guidelines, further optimize algorithms, conduct large-scale, rigorous randomized clinical trials, and improve privacy and security mechanisms to ensure the safety, reliability, and sustainability of AI applications in this field.