AI in health is often seen as objective, but this neutrality can mask deep inequities. This talk explores how data-driven systems can overlook vulnerable populations, misclassify risk, and reinforce disparities, revealing the hidden cost of using “objective” AI in population health.