A healthcare AI system predicts patient readmission risk. An audit reveals the model performs significantly worse for patients from low-income zip codes compared to high-income ones, though both groups are represented in the training data. What type of bias is MOST likely present?

C is correct because measurement bias arises when features (such as zip code as a proxy for healthcare access) contain systematic errors that differ across groups. Low-income zip codes may have lower data quality or proxies that behave differently than assumed. A is less likely since both groups are represented. B (label bias) would show accuracy parity but label correctness issues. D does not apply since low-income patients are within scope.