---
sourceDocument: Australia IT Operations Management
sourceDocumentLink: https://www.servicenow.com/docs/r/australia/it-operations-management

 Release :

    - australia

ft:locale :

    - en-US

ft:publication_title :

    - Australia IT Operations Management

ft:clusterId :

    - itom

bundleId :

    - itom

workflow :

    - Technology


---

# Create AI Control Tower (AICT) policies

# Create AI Control Tower (AICT) policies {#ariaid-title1}

Release version: Australia  
Updated October 2, 2026  
![](https://www.servicenow.com/docs/portal-asset/ico-clock) 1 minute to read  
Create AICT policies to block personnel from using unauthorized AI tools. AICT polices are synchronized with the Agent Client Collector, which invokes the policies that run.

## Before you begin

You must configure ACC with an ICS (mid-less) architecture to enable using an AICT policies. For details on configuring ACC without a MID Server, see [Configuring MID-less Agent Client Collector](https://www.servicenow.com/docs/XzdpoJFIS~3grGL0dok96w "Configure MID-less Agent Client Collector to enable sending information through the cloud. Sending information through the cloud allows the MID Server to be used for more persistent resources.").

Ensure that ITOM Shadow AI governance is installed in your enviornment.

Role required: agent_client_collector_admin

## Procedure

1. Navigate to AllAI Control TowerAI Assets.
2. In the filter on the left side of the page, expand Govern and select Policies.
3. Select Create policy.
4. Select a policy type in the Choose a policy type field.  
   * A threat response: Takes an agent offline when a security threat is detected.
   * An explicit block: Blocks a user or department from using an AI agent, model, or domain.
   {#create-aict-policies__ul_vmb_ksr_tkc}
5. Configure the relevant parameters, based on the selected policy type.  
   * A threat response:
     1. Select what will happen when a threat is detected.
     2. In the Which agents cell, select the conditions by which the detected threat takes the agent offline.
     3. In the sensitivity field, select whether the agent is to be taken offline on every detection, or only above a specified threshold.
     4. Select a follow-up action in the Do this cell, after the agent is taken offline:
        * Create a ticket: Opens a ticket in the indicated group's queue.
        * Notify: Sends a notification to the user.
        {#create-aict-policies__ul_nfm_ztr_tkc}
     5. Select the Publish button to publish the policy.
     {#create-aict-policies__ol_y2v_zsr_tkc}
   * An explicit block:
     1. Select whether the block is to apply to a user or a department in the Who field.
     2. Select what is being blocked in the What field, either an AI agent, a model, or a domain.

        When selecting an AI agent, configure a filter for the field values that create the
        block, as needed.
     3. Select a follow-up action of Notify to send a notification to the user, in the Add follow-up drop down.
     4. Select the Publish button to publish the policy.
     {#create-aict-policies__ol_q14_btr_tkc}
   {#create-aict-policies__ul_qhg_ysr_tkc}

## Result

* The action to be taken appears in the Policy summary cell.
* The policy is synchronized with the Agent Client Collector policies, and appears in the list of Agent Client Collector policies. Agent Client Collector invokes the policy and takes the indicated action.
{#create-aict-policies__ul_a15_5wr_tkc}

*[\>]: and then