AI RESEARCH
When Background Matters: Breaking Medical Vision Language Models by Transferable Attack
arXiv CS.CV
•
ArXi:2604.17318v1 Announce Type: new Vision-Language Models (VLMs) are increasingly used in clinical diagnostics, yet their robustness to adversarial attacks remains largely unexplored, posing serious risks. Existing medical attacks focus on secondary objectives such as model stealing or adversarial fine-tuning, while transferable attacks from natural images