Artificial intelligence in healthcare: Real-world integration, cybersecurity risks and challenges

Abstract

Advancements in artificial intelligence (AI) continue to drive innovation across industry, society, and the economy, including the healthcare sector. Since the initial adoption of AI technologies, healthcare providers have increasingly integrated these tools to enhance accuracy, efficiency, and decision support in diagnosis, treatment, and individualized patient care workflows. Applications such as medical imaging analysis, early disease detection, and administrative optimization have moved from trials into real-world deployments, offering valuable insights into both the benefits and emerging risks of AI in clinical settings. This paper expands on a previous study and presents an updated assessment of AI integration in healthcare, with a specific focus on cybersecurity risks and real-world developments since 2023. Using a selective literature review combined with examination of recent deployments and evolving regulatory frameworks, the paper discusses how foundational cybersecurity vulnerabilities, already present in healthcare infrastructure, are intensified by AI-specific challenges, including data governance weaknesses, model trustworthiness, and the risk of biased or manipulated outputs. The research also highlights how uneven regulatory landscapes across countries create inconsistencies in AI oversight, increasing exposure to cyber threats. Identified examples, such as the integration of AI in radiology, illustrate how these trends manifest in practice. Building on these findings, the paper discusses the current progress in ensuring a secure AI adoption in healthcare and outlines recommendations to mitigate these complex risks and promote safe, ethical, and effective use.