Privacy-preserving AI in healthcare: a comprehensive survey of techniques and challenges

The integration of artificial intelligence (AI) into healthcare systems has unlocked transformative potential in areas such as medical diagnostics, personalized treatment, and patient monitoring. However, the sensitive nature of medical data introduces significant privacy and security concerns. This survey presents a comprehensive review of privacy-preserving techniques tailored for AI-driven healthcare applications. We begin by identifying and analyzing the key challenges involved in safeguarding personal health information within intelligent systems. Next, we propose a structured classification of privacy-preserving methods into four main categories: cryptographic approaches, decentralized learning methods, perturbation-based techniques, and hybrid models. For each category, we outline the underlying principles, assess their strengths and limitations, and examine their applicability to real-world healthcare scenarios. Furthermore, we explore practical solutions that integrate multiple techniques and compare them in terms of accuracy, scalability, computational cost, and data utility. Finally, we discuss open issues and future research directions, including the trade-offs between privacy and model performance, regulatory compliance, and the integration of generative AI and large language models (LLMs) into privacy-aware frameworks. This study aims to guide researchers and practitioners in designing secure, efficient, and privacy-preserving AI systems for healthcare.