AI is increasingly used in healthcare, from disease screening to treatment recommendations. But what happens when the data used to train these systems reflect existing inequalities? In its first newsletter, AEQUITAS shares early findings on gender and racial bias in biomedical AI and outlines what the project is doing to help healthcare professionals and civil society organisations address these risks.

The 1st newsletter introduces the AEQUITAS Bias Report, which examines how bias can arise at different stages of AI development and use in three areas: cardiovascular disease, diabetes and depression. One example comes from cardiovascular care. When women are referred for diagnostic tests less often than men, an AI system trained on those records may learn patterns shaped by unequal access to testing.

The newsletter also previews what’s next for AEQUITAS: an interactive database documenting examples of bias, a practical AI Regulatory Model grounded in the EU Charter of Fundamental Rights, and training workshops for healthcare professionals across Europe.