AI RESEARCH
Making MLLMs Blind: Adversarial Smuggling Attacks in MLLM Content Moderation
arXiv CS.CV
•
ArXi:2604.06950v1 Announce Type: new Multimodal Large Language Models (MLLMs) are increasingly being deployed as automated content moderators. Within this landscape, we uncover a critical threat: Adversarial Smuggling Attacks. Unlike adversarial perturbations (for misclassification) and adversarial jailbreaks (for harmful output generation), adversarial smuggling exploits the Human-AI capability gap. It encodes harmful content into human-readable visual formats that remain AI-unreadable, thereby evading automated detection and enabling the dissemination of harmful content.