Douglas Teodoro, Assistant Professor in the Department of Radiology and Medical Informatics at the Faculty of Medicine, and Caroline Samer, Associate Professor and Director of the Department of Anaesthesiology, Pharmacology, Intensive Care and Emergency Medicine at the Faculty of Medicine, group leader at the Section of Pharmaceutical Sciences of the Faculty of Sciences, and Head of the Division of Clinical Pharmacology and Toxicology at the HUG, will lead two major parts of the European project AIM-SAFE, which brings together a consortium of 25 institutions across 14 countries. The project aims to develop AI-powered clinical decision support tools based on generative artificial intelligence to improve the safety of drug treatments — particularly for patients on multiple medications and in oncology. Funded by Horizon Europe within the GenAI4EU initiative, which seeks to foster "Made in Europe" generative AI projects, the total budget amounts to €16.7 million, of which €1.73 million is allocated to the Geneva teams.

Designed for direct deployment in hospital settings, AIM-SAFE relies on several specialised AI agents capable of analysing complex biomedical data to predict adverse drug events, identify deprescribing opportunities, integrate pharmacogenomics into therapeutic decision-making, and optimise dosing regimens. The project was initiated and structured by the Data Science for Digital Health (DS4DH) group, led by Douglas Teodoro, a specialist in data science applied to health. He will lead the development of personalised predictive models for prescribers and clinical pharmacologists. Caroline Samer, a specialist in pharmacogenomics and drug safety, will lead alongside Dr Aurélien Simona, a senior clinician at the HUG, the clinical validation of the system across European hospitals — the critical step to test these tools under real-world conditions.

By fostering more personalised and safer medicine, AIM-SAFE could help reduce adverse drug events, improve the quality of care and lower healthcare costs. The project also aims to develop a reliable, explainable and privacy-preserving European medical AI, fully compliant with European regulations (GDPR, AI Act, European Health Data Space).