HijackRAG: Hijacking Attacks against Retrieval-Augmented Large Language Models

Open in new window