Health Equity & Algorithmic Bias

Analysis and reporting on disparities, bias risks, and equity implications of AI in healthcare. Covers documented cases of algorithmic bias (e.g., pulse oximeter inaccuracy in darker-skinned patients, sepsis models trained on non-diverse populations, dermatology AI underperformance on underrepresented skin tones), structural factors that produce biased training data, regulatory and policy responses, and frameworks for bias auditing and mitigation. Content cites primary research and policy sources including KFF, WHO equity guidance, and peer-reviewed literature. This group is a distinct editorial commitment, not a tag or subcategory, because equity and bias considerations cut across every clinical domain and technology type and require dedicated, sustained coverage. Serves clinicians, policy professionals, health system leaders, and researchers who recognize that AI performance is not population-neutral.