---
sourceDocument: Australia Build workflows
sourceDocumentLink: https://servicenow-prod.fluidtopics.net/r/build-workflows

 Release :

    - australia

ft:locale :

    - en-US

ft:publication_title :

    - Australia Build workflows

ft:clusterId :

    - crworkflow

bundleId :

    - crworkflow

workflow :

    - Creator


---

# Use AI agent as an activity

# Use AI agent as an activity {#ariaid-title1}

* Release version: Australia
* 
* Updated March 12, 2026
* 
* ![](https://www.servicenow.com/docs/portal-asset/ico-clock) 3 minutes to read

Use an existing AI agent as an activity in your playbook to automate tasks. The AI agent gathers the required context, performs the work, and produces outputs that subsequent activities can consume.

## Before you begin

Make sure that ServiceNow Otto® AI agents plugin is installed and you have AI agents available in your instance. For information about installing the ServiceNow Otto® AI agents plugin, see [Install ServiceNow Otto AI Agents](https://www.servicenow.com/docs/access?context=install-ai-agents-plugins&version=australia&pubname=australia-intelligent-experiences&ft:locale=en-US). For information about creating an AI agent, see [Create an AI agent](https://www.servicenow.com/docs/access?context=configure-next-best-action-agent&version=australia&pubname=australia-intelligent-experiences&ft:locale=en-US).

Role required: playbook.admin or pd_author.

## About this task

When you add an AI agent to the activity, you select the agent, provide an objective for the agent to complete, and define the expected outputs. You can also define whether you want to run the agent on behalf of the user who
triggered the playbook, or the current user. You can interact with the agent from within the playbook, without opening the ServiceNow Otto® panel.

## Procedure

1. Navigate to AllProcess automationWorkflow Studio.
2. Select the playbook where you want to add the Use an AI agent activity.
3. In a new or an existing stage, select Add an activity.  
4. From the list of activities, under Common Activities, select Use an AI agent.
5. In the Automation tab, configure the AI agent.  
   {#ai-agent-as-activity__table_slj_c1t_33c__entry__2}

   | Field | Description |
   |-|-|
   | AI agent | The AI agent that you want to add to the activity. |
   | Objective | Text directions for the AI agent. You can use a combination of plain text and data pill picker. Provide a descriptive instruction for the AI agent about what do you want to achieve. For example, <kbd class="ph userinput">Summarize the incident (Parent record-&gt;Incident-&gt; Number) and suggest the next steps to resolve the incident.</kbd>. |
   | Run as | The user on behalf of whom you want the agent to run. * Active user: The AI agent performs the actions on behalf of the active user that starts the conversation. * User who triggered the playbook: The AI agent performs the actions on behalf of the user that triggered the playbook. * User who completed a prior activity: The AI agent performs the actions on behalf of the user that completed a previous activity. {#ai-agent-as-activity__ul_o2n_l3t_33c} |
   | Expected outputs | Outputs that you want the AI agent to generate. |
   [Table 1. Use AI agent properties]

   {#ai-agent-as-activity__table_slj_c1t_33c}
6. **Optional:** Select Show additional options to configure the optional properties.  
   {#ai-agent-as-activity__table_rsc_pht_33c__entry__2}

   |   |   |
   |-|-|
   | Assigned to | An optional field where you can specify a user who can complete, skip, or restart the activity when the agent is working. This field is visible only when you select Show additional options. |
   | Assignment group | An optional field where you can specify a user group who can complete, skip, or restart the activity when the agent is working. This field is visible only when you select Show additional options. |
   | Conversation label | A name to identify the conversation. |
   [Table 2. Optional properties]

   {#ai-agent-as-activity__table_rsc_pht_33c}
7. Select Save and close.

## Using AI agent in a CSM playbook {#ai-agent-as-activity__example_b15_nnt_33c}

In the following example, Alex, a CSM/FSM Workspace user, is working on a customer complaint about a product issue. The playbook uses the Complaint Case Research Plan Agent in the research stage to analyze the complaint and
create relevant tasks for the next steps. The AI agent is configured to run as the active user in this playbook.

Alex performs all interactions with the AI agent within the playbook, without opening the Now Assist panel.

1. When the playbook reaches the research stage, Alex selects Start Now Assist to run the AI agent.

2. Alex can view the tasks that the AI agent is performing.

3. After analyzing the case and similar cases, the AI agent recommends adding four tasks and prompts Alex for confirmation. Alex can remove any of the recommended tasks or select Confirm for the AI agent to create the tasks.

4. After receiving the confirmation, the AI agent creates the tasks and marks the activity as complete.

{#ai-agent-as-activity__ol_zs4_gpb_j3c}

*[\>]: and then