AI for Science & Society

Causal ai for medicine safety

Quantifying treatment risks by connecting causal inference, machine learning, and mechanistic evidence.

This programme develops causal-inference and AI/ML pipelines for medicine safety. A central application is the quantification of antibiotic-associated risk, integrating observational data, biological mechanisms, and explicit assumptions about treatment, exposure, and outcome.

The broader aim is to move from prediction alone toward models that support transparent scientific and regulatory decisions.