---
sourceDocument: Brazil Enable AI
sourceDocumentLink: https://www.servicenow.com/docs/r/intelligent-experiences

 Release :

    - brazil

ft:locale :

    - en-US

ft:publication_title :

    - Brazil Enable AI

ft:clusterId :

    - platai

bundleId :

    - platai

workflow :

    - Platform


---

# Define expected behavior

# Define the expected behavior of an AI agent {#ariaid-title1}

Release version: Brazil  
Updated October 23, 2026  
![](https://www.servicenow.com/docs/portal-asset/ico-clock) 2 minutes to read  
In the guided setup for an AI agent, write a clear description defining your AI agent, its role, and the instructions it should perform to accomplish tasks.

## Before you begin

You must either create an AI agent or edit an existing one to get to this step in the guided setep. To learn how to get started creating an AI agent, see [Create an AI agent](https://www.servicenow.com/docs/HrWjLz52~GRSkMnKL44GFQ "Create an AI agent in AI Agent Studio to solve problems for your users and coordinate with other AI agents in agentic workflows."). Otherwise, navigate to the AI Agent Studio and select the AI agent you want to define to enter the guided setup.

The name, description, modality, AI agent role, and instructions are all required. If the agent was drafted in a conversation, or created from an opportunity, these fields are pre-filled as a first draft. Treat pre-filled content as generated text to be reviewed, not as confirmed configuration.

Role required: sn_aia.admin

## About this task

The first step of the guided setup enables you to define the fundamentals of the AI agent. The description, AI agent role, and instructions fields are used by the AI agent Orchestrator to understand how to use the AI agent, by itself or as
part of an agentic workflow. Descriptions, AI agent roles, and instructions should be clear and well-defined. For guidelines for writing these fields, see [Writing effectively for agentic AI](https://www.servicenow.com/docs/6~S4lfOX4tdkrC93vMC37Q "By following some general guidelines for creating AI agents and agentic workflows, you can create clear and effective instructions that help maximize their efficiency and effectiveness."). For an example AI agent, see [Example AI agent](https://www.servicenow.com/docs/LaSeg_JO1ngw557Wp5NSQw "Use the example AI agent with clear name, description, AI agent role, and list of steps fields to use as a guide when creating your own AI agents.").

## Procedure

1. Verify the name and description summarize the main intention of the AI agent.  
   If the name and description aren't clear or correct, the Orchestrator won't know when or how to use the AI agent.
2. Check that the modality of the AI agent is correct.  
   The Chat modality allows users to invoke the AI agent in Virtual Agent or as part of an agentic workflow in the ServiceNow Otto panel.

   An agent has one modality. Chat and Voice cannot both be applied to the same agent. If you need both channels, build one agent per modality. Selecting Voice also requires that a voice assistant already exists in Assistant Designer.

   See [Create an AI voice agent](https://www.servicenow.com/docs/JdW2~dsQDNVr9eBosBrRTQ "Create an AI voice agent in the AI Agent Studio to resolve cases, incidents, or tasks through the phone channel.") for more information about creating AI agents with the Voice modality.


3. Review and refine instructions.  
   The instructions are the most important element of the AI agent. While the name and description guide the AI agent Orchestrator to know when to invoke an AI agent, the instructions tell the Orchestrator what it does and how it will do tasks.

   You can create different versions of the AI agent instructions. After [testing an AI agent](https://www.servicenow.com/docs/9vXwu1HOXxrPfmCgyw5Hrg "Manually test an AI agent or agentic workflow to verify it functions as defined and achieves the desired objectives.") or [evaluating an AI agent](https://www.servicenow.com/docs/ZoN4KUfQyZIOIoMfw__CAw "Evaluate your agentic AI asset's performance against defined scenarios to identify issues and opportunities for improvement."), you can adjust the instructions and compare their performance. See [Version control for agentic AI](https://www.servicenow.com/docs/X7glerPpTS84G2oAoWv63g "Version control enables you to track changes made to instructions sent to the LLM for AI agents and agentic workflows.") to learn more.
4. **Optional:** Confirm the AI agent's expertise.  
   The difference between the description and the agent role is that the agent role provides more detail about the function that the AI agent serves within a greater context. The greater context may be as part of an agentic workflow, or it could be the specific function it serves in another type of work process.


## Result

The fundamentals of your AI agent are defined so that the AI agent Orchestrator can use it.

## What to do next

Scroll down to the next section of the guided setup, Tools, to [add tools and information](https://www.servicenow.com/docs/4zuFIUomTGbD7imZq8ovlQ "Add a tool to an AI agent to enable different functionalities and help your AI agents achieve their objectives.") your AI agent can use.

