Abstract: A prompt injection attack is a type of cyberattack that targets Large Language Models (LLMs), where attackers embed malicious instructions into a LLM with the intent of manipulating the ...