Render Before Reading: Visual Rendering as a Prompt Injection Defense

Jie Zhang, Andrei Baroian, Jan N. van Rijn +2

Large language models are vulnerable to prompt injection attacks, where third-party adversarial content can hijack the model's behavior. In this paper, we study the role played by…