How to stop Bot from responding to unethical questions

AI for Process
1

Hi,
I have integrated the Azure Open AI and using Agent node, prompt nodes, dialog gpt. I want to stop my bot from responding to unethical questions like “how to reset my friends passwords” and give a fallback message like you are not authorized to do this task using the AI. Please help me to achieve this in my bot.

1 Like
2

Hey Gayapu Srinivasa Reddy,
Welcome to the Kore.ai Community!
To answer your question, there’s something called Guardrails. They enable responsible and ethical AI practices by allowing platform users to easily enable/disable rules and configure settings for different features using LLMs.
To navigate you correctly to the Guardrails option on the platform:

  1. Go to Generative AI Tools.
  2. Upon opening it, you’ll see Safeguards – click on that.
  3. You’ll then see an option called Guardrails.
  4. Click on Guardrails, and you will be able to see all the guardrail features.

Enabling these settings can help address your question, and hopefully, this will guide you towards your solution.

For more detailed information, please refer to this link:
https://docs.kore.ai/xo/generative-ai-tools/guardrails/

Thank you,
Jaswanth Narne
Kore.ai Community Team