Muhammad Umer

VisGuard-Ur: A Proof-of-Concept Study on Typographic Jailbreak and Prompt-Injection Attacks Against Urdu-Aware Vision-Language Models

Vision-language models (VLMs) are increasingly deployed as document- and image-understanding agents, yet published safety evaluations of these systems are almost exclusively conducted in English plain text. This leaves two attack surfaces largely unexamined for low-resource languages: (1) adversarial instructions embed...