Read news on harmful prompts with our app.
Read more in the app
LLMs respond differently to harmful prompts when AI watermarking is used