Skip to main content
. Author manuscript; available in PMC: 2024 Jul 14.
Published in final edited form as: Science. 2023 Jul 13;381(6654):150–152. doi: 10.1126/science.adh2713

How a “fair” algorithm can result in biased outcomes.

How a “fair” algorithm can result in biased outcomes

A clinician and patient interact with artificial intelligence (AI)–based decision support that provides information about, for example, the likelihood of a diagnosis, utility of a treatment, or a prognosis. Even if the algorithm is unbiased, clinician-, patient-, and social-level factors can influence how the recommendations are interpreted and implemented, which can result in latent treatment bias.