Artificial intelligence (AI) is increasingly integrated into healthcare systems across Europe, supporting diagnosis, clinical decision-making, treatment planning, and patient risk prediction. These technologies offer considerable opportunities for improving healthcare efficiency and outcomes. At the same time, growing evidence shows that AI systems can reproduce or amplify existing inequalities in healthcare

A well-documented example is an exploratory study in the Journal of Medical Internet Research which found sex-based performance disparities in machine-learning algorithms used for cardiac disease prediction, in particular that female patients were consistently underrepresented in training datasets and that models showed higher false-negative rates for women, demonstrating how biased data can contribute to inequitable diagnostic performance.

Addressing these bias risks requires regulatory models that translate international and European ethical and legal principles into concrete governance mechanisms that healthcare institutions can implement in practice.

This policy document proposes a regulatory model for biomedical AI governance which aligns with the EU acquis, including the EU Fundamental Rights Charter, and is designed for public healthcare systems and civil society organisations (CSOs).

DOWNLOAD

To έγγραφο It is part of the project “AEQUITAS – Preparing Civil Society Organisations and the public healthcare sector to address gender and racial biases that may arise from the widespread use of Artificial Intelligence, with the aim of protecting and promoting fundamental rights.” The project is implemented by a consortium of nine organisations: Center for the Study of Democracy (CSD) (coordinator), Diotima Centre, Klinikum der Universität zu Köln (UKK), Innovation Hive (INNOHIVE), Citizens in Power (CIP), Women’s Issues Information Centre (WIIC), Lobby Europeo de Mujeres en España (LEM), TIA Formazione Internazionale APS, and Health Citizens European Institute.
It is funded by the European Union under the CERV-2024-CHAR programme. Visit the project page here.