As healthcare organizations increasingly adopt AI-driven clinical tools, new questions are emerging about how patient data is used and protected. This session explores the concept of inference contamination, where AI recommendations are shaped by patterns learned from many patients’ data, even without direct access to individual records. Attendees will examine the limitations of traditional HIPAA frameworks in addressing these indirect data influences and the unique privacy challenges introduced by machine learning. The presentation highlights gaps in current de-identification practices and introduces emerging techniques such as differential privacy. Participants will leave with practical strategies to guide policy development, strengthen data governance, and redefine privacy in an AI-enabled healthcare environment.

This session offers 1 AHIMA CEU.