From networks of data to networks of care in clinical medicine: this is not artificial intelligence.

The increasing availability of clinical, omic, and imaging data in clinical medicine opens unprecedented opportunities to uncover hidden patterns and mechanistic insights into disease. Available tools augment clinical reasoning, equipping clinicians with new layers of understanding to support personalized treatment and transparent risk assessment. More tools are being implemented and will soon enter such a challenging field of study and application. Graph- and sequence-based Artificial Intelligence (AI) operating on these networks underpins the vision of precision medicine, enabling early diagnosis, disease staging, treatment selection, and transparent risk stratification. Crucially, eXplainable AI (XAI) methods attribute predictions to specific nodes and sub-networks (e.g., genes, proteins, metabolites, microbes, clinical features), aligning model outputs with clinical reasoning and regulatory expectations. Rather than replacing clinicians, these tools augment clinical attentiveness and insight by tailoring decisions to each patient's molecular and clinical profile. By modelling complexity in an interpretable way, AI and network science convert performance into practice. Internists must be prepared to face this rapidly growing technological challenge. Here, we explore how network-based modelling and XAI can enhance clinical reasoning by revealing emergent behaviors, stratification patterns, and novel interactions across biological systems.