Artificial Intelligence in Psychiatric Inpatient Care: Advancing Diagnostics, Personalized Treatment, and Ethical Integration

Abstract

Artificial intelligence (AI) is transforming mental inpatient care by improving diagnostic precision, facilitating individualized therapy, and optimizing hospital operations. This scoping review aggregated findings from 24 empirical studies published between 2015 and 2025 to assess the application of AI technologies—such as machine learning, natural language processing, deep learning, digital phenotyping, and conversational agents—in inpatient psychiatric environments. Findings demonstrate that AI enhances the early identification of relapse and suicide risk, facilitates personalized therapy via decision-support systems and chatbots, and bolsters patient monitoring using sensor-based technology. AI enhances operational efficiency by optimizing bed allocation, personnel scheduling, and clinical documentation, hence alleviating administrative burdens. Nonetheless, considerable obstacles persist, including algorithmic bias, privacy issues, clinical opposition, and legal uncertainty. This study offers the Three-Pillar Model for Responsible AI Integration, highlighting therapeutic augmentation, ethical safeguards, and operational governance as fundamental concepts. The analysis highlights the dual nature of AI in psychiatry: its revolutionary promise alongside ethical and implementation challenges. Future investigations should prioritize longitudinal validation, resource-constrained environments, interpretability, and the creation of inclusive datasets. By incorporating transparency, fairness, and human-centered design, AI can enhance mental inpatient treatment to be technologically advanced, equitable, trustworthy, and compassionate.