Health Affairs April 25, 2024
Artificial intelligence (AI) (such as machine learning) has been used in health care for decades, but the sector is currently seeing a new wave of excitement for its potential to improve patient outcomes, cost-effectiveness, and efficiencies in care provision. Creating new hope is AI’s potential to redress historic inequities. For example, it could aid the identification of biases and provide underserved groups with access to personalized medicine. However, of growing concern is the danger that AI-enabled algorithms used in health care could perpetuate discriminatory bias, resulting in worsening disparities—especially in already vulnerable populations. The field is only just learning how to reduce such biases, arguably at a much slower pace than technology advancement and deployment. Garba-Sani’s research over the past...