THM Writeup | HealthGPT

Title: HealthGPT | Category: AI/LLM / Prompt Injection | Difficulty: Easy Introduction In the world of AI security, ”Refusal Leakage” is a subtle but effective vulnerability. It occurs when a Large L