Introduction: The rapid integration of Artificial Intelligence (AI) into healthcare promises transformative advancements in diagnosis, personalized treatments, and operational efficiency. However, this progress is inherently linked to the processing of vast, sensitive patient datasets, raising significant concerns about data privacy and security. Objectives: This literature review systematically examines the complex landscape of privacy and data anonymization within AI-driven healthcare, highlighting the critical risks of re-identification from highly sensitive biomedical data, including facial images and genetic information. Methods: This integrative literature review (2020-2025), based on studies from PubMed, Scopus and Web of Science databases, investigated safeguarding Privacy in the Age of AI, focusing on ethical and technical challenges in healthcare data anonymization. Results: Although research on Generative Adversarial Networks (GANs) in healthcare AI is expanding, there are still obstacles to overcome. The amount of computing power needed to train GANs is also substantial. Federated learning is not impervious to all types of privacy violations, even though it keeps data on local devices. Explainable AI (XAI) and privacy, Privacy-Preserving Machine Learning (PPML) frameworks, synthetic data generation outside of GANs, hardware-based privacy solutions, and ethical AI governance and auditing are some of the new developments and research gaps in privacy-preserving AI in healthcare. Conclusions: It is concluded that data privacy and patient confidence are two issues that artificial intelligence in healthcare raises. Both contemporary and traditional anonymization techniques are essential, but data utility and individual privacy must be balanced. Strong regulation and ethical cooperation are crucial.
Lima et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: