Announcement_30_1
🧬 Our work on secure and trustworthy healthcare AI has been published in Nature Scientific Reports!
Our paper titled, “Secure yet fragile: adversarial vulnerabilities of federated vision–language models in medical AI”</span> examines federated Vision-Language Models for medical AI and demonstrates how adversarial perturbations introduced at one participating site can propagate through collaborative training and substantially degrade the shared model’s performance.
🔬 Read the Paper in Scientific Reports