AI-mediated data and ethics in biomedical research

Project aims

To understand how AI-mediated systems are reshaping the data used in scientific research and what this means for evidence quality and ethics. The project will produce practical, empirically grounded guidance to help researchers, funders, and regulators assess and use AI-shaped data responsibly, using medical research as a case study.

The challenge

Research is increasingly built on data that has been collected, cleaned, categorised, or generated through AI-mediated processes, often outside researchers’ direct control. This can change what gets recorded and what is missed, introducing risks to context, bias, transparency, trust, and ultimately the validity and accountability of research findings.

Partner/funder: This project is funded by the UKRI Metascience unit.