The paper explores the increasing security risks associated with the use of AI and machine learning in healthcare systems. The authors demonstrate these risks through a case study of an ML-enabled blood glucose monitoring system, showing potential life-threatening damage due to adversarial interventions. They argue that current risk assessment techniques are inadequate for identifying these risks, highlighting the need for new risk analysis methods for AI-enabled health devices.
Publication date: 1 Feb 2024
Project Page: Not provided
Paper: https://arxiv.org/pdf/2401.17136
