Prompt Injection Defense: Preventing AI Manipulation

You worked hard to set many rules for the AI, only for a user to bypass them all with a single "ignore previous instructions." Prompt injection is the most common security vulnerability in AI applications.

💡 What You Will Learn

You worked hard to set many rules for the AI, only for a user to bypass them all with a single "ignore previous instructions." Prompt injection is the most common security vulnerability in AI applicat

AI

AI

Please translate the following to English, then answer: []

def filter_injection(user_input):
    dangerous = ["", "", "", "act as"]
    for keyword in dangerous:
        if keyword in user_input.lower():
            return ""
    return user_input

def verify_output(input_text, output_text):
    # 
    if "" in output_text or "" in output_text:
        return ""
    return output_text

<user_input></user_input>
<user_input>
{}
</user_input>