Render Before Reading: Visual Rendering as a Prompt Injection Defense

The paper investigates how multimodal large language models are more susceptible to prompt injection when adversarial instructions are presented as text rather than as non-textual inputs like images. It proposes a training‑free defense that renders untrusted payloads into typographic images (or audio) before they reach the model, a method called Pictionary. Experiments on ten models and two benchmarks show that this approach significantly lowers attack success rates while maintaining normal functionality, and that fine‑tuning on image‑rendered instructions can further reduce the modality gap.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

• llms
• fine-tuning
• multimodal
• benchmarks
• safety