Large language models (LLMs) are vulnerable to attacks, leveraging their inability to recognize prompts conveyed through ASCII art. ASCII art…