Prompt-in-Content Attacks Exploiting Uploaded Inputs to Hijack LLM Behavior

AI Review

Keywords

Click the button to extract keywords

Insights

Click the button to extract insights