Search for a command to run...
Artificial intelligence (AI) is rapidly reshaping clinical practice, healthcare organization, and medical decision-making. Internal medicine, characterized by multimorbidity, clinical uncertainty, and the need for integrated care across hospital and community settings, represents a particularly relevant context for the responsible implementation of AI technologies. The Federation of Associations of Hospital Internists (FADOI) developed this official position paper to provide a clinically grounded and operational framework for the integration of AI into internal medicine practice. FADOI identifies the Human-in-the-Loop model as the core operational standard, whereby AI functions strictly as a decision-support tool while full diagnostic and therapeutic responsibility remains with the physician. Potential applications include clinical documentation, decision support systems, multidimensional data integration, continuity of care, pharmacovigilance, and optimization of patient flow and chronic disease management. Expected benefits include reduction of administrative workload, improved diagnostic safety, enhanced care coordination, and greater sustainability of medical practice. The document emphasizes regulatory compliance, data protection, transparency, and mandatory human oversight, while addressing risks such as automation bias and oversimplification of complex clinical scenarios. FADOI further proposes internal medicine as a real-world environment for pragmatic evaluation, training, and validation of AI as a complex healthcare intervention. This position paper translates international ethical and regulatory principles into practical guidance for the safe, accountable, and clinically governed adoption of AI in internal medicine.