​
Introduction

Deciding on the right policies your agent must follow is an important part of the agent building process. Agents are going to represent you, your brand or your customers. You must have full confidence in how they will deal with specific situations.

​
What is policy?

Policy is like the Agent Guardrails and safeguards applied to each AI agent, allowing for control over AI interaction.

For example, you can control:

  • Whether it can stray from its knowledge context
  • Whether it can talk about safeguarding issues
  • If somebody mentions safeguard issues, what contact it should provide or what message it should send
  • Whether it can ask questions back
  • Whether it can discuss legal or financial matters

There are several key benefits of using Mindsets policy controls with agents:

  • Trust: Builds confidence among your customers and team that agents will deal with sensitive issues or ethical matters in the correct way.
  • Reliability: Ensures AI agents operate within predefined boundaries, reducing errors and maintaining quality outputs.