The AEQUITAS project is pleased to announce the launch of the AEQUITAS Database, a major project outcome developed under Task T2.3. The database has been designed as a structured and accessible platform for documenting and exploring gender and racial biases in biomedical Artificial Intelligence (AI).
As AI technologies are increasingly integrated into healthcare systems, concerns regarding fairness, transparency, and discrimination have become more prominent. AI-driven tools used in diagnosis, risk prediction, and clinical decision-making may unintentionally reproduce or amplify existing inequalities when trained on biased or non-representative datasets. The AEQUITAS Database addresses this challenge by collecting and organizing evidence related to such biases in a user-friendly and searchable environment.
Supporting More Inclusive Healthcare
The database focuses particularly on three major disease areas:
- Cardiovascular disease
- Depression
- Diabetes
These areas were selected due to their high societal impact and the documented disparities that affect different gender and racial groups. Through the database, users can access curated entries containing information such as:
- Study title and source
- Disease area
- Type of bias identified
- Application scope of the AI system
- Description of the findings
- Population groups affected
The platform also includes filtering and search functionalities, allowing users to conduct targeted exploration and bias analysis.
A Resource for Multiple Stakeholders
The AEQUITAS Database has been developed to support a wide range of stakeholders, including:
- Healthcare professionals
- Civil society organizations
- Researchers
- Policymakers
Healthcare professionals can use the platform to better understand potential biases in AI-assisted healthcare tools and support more equitable clinical decision-making. Civil society organizations may use the database for awareness raising and advocacy activities related to the protection of fundamental rights.
A Collaborative and European Effort
The database currently includes 100 curated entries derived from scientific publications, reports, and additional sources identified by project partners. In addition to English-language literature, multilingual sources from the national languages of consortium partners have also been incorporated, ensuring broader representation and contextual diversity. The database is publicly accessible through the official AEQUITAS project website and can be accessed via the “Results” section: https://aequitasproject.eu/results/ . Users are required to complete a registration process before obtaining access to the platform, ensuring responsible use and compliance with data protection and ethical standards.
Promoting Trustworthy AI in Healthcare
The launch of the AEQUITAS Database represents an important step towards promoting fair, transparent, and trustworthy AI systems in healthcare. By transforming research findings into a practical and accessible digital tool, the AEQUITAS project contributes to reducing inequalities and supporting more inclusive healthcare systems across Europe.
