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

 Release :

    - zurich

ft:locale :

    - en-US

ft:publication_title :

    - Zurich Enable AI

ft:clusterId :

    - platai

bundleId :

    - platai

workflow :

    - Platform


---

# Managing your AI asset inventory

# Managing your AI asset inventory {#ariaid-title1}

Release version: Zurich  
Updated August 5, 2026  
![](https://www.servicenow.com/docs/portal-asset/ico-clock) 5 minutes to read
Summarize  
![AI sparkle icon](https://servicenow.com/docs/portal-asset/ai-sparkle-icon) Summarized using AI  
This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.  

## Summary of Managing your AI asset inventory

The AI asset inventory in AI Control Tower is a centralized record that tracks all AI assets across your organization, including AI systems, models, prompts, and datasets.
This inventory enables you to control governance, monitoring, and risk workflows by managing asset participation and providing visibility into asset details and statuses.
Assets can be discovered automatically or added manually to maintain a comprehensive and accurate inventory.
Show full answer Show less  

## Key Features

* **Comprehensive Asset Tracking:** Automatically discover and manually add AI assets such as AI systems (Agentic AI, Generative AI, Classic AI), AI models, AI datasets, and AI prompts to monitor their lifecycle, risk, and governance status.
* **Governance and Management Control:** Set the managed status of assets to include or exclude them from governance workflows, risk classification, value assessment, security, and privacy controls.
* **Filtering and Refinement:** Filter assets by type, risk classification, lifecycle state, asset status, and management status to focus on relevant assets that require attention.
* **Role-Based Access:** Access to the AI asset inventory requires the AI steward or AI system owner roles, ensuring proper governance oversight.
* **Bulk Management Automation:** Use automation rules to classify and manage large numbers of AI assets based on defined conditions, streamlining governance processes.
* **Evaluation and Monitoring:** Enable evaluation scoring to detect quality and safety regressions in managed assets, helping maintain AI system reliability and compliance.
* **Duplicate Asset Review:** Identify and review potential duplicate AI assets to avoid redundancy and ensure efficient resource use.
* **Deactivation Capability:** Deactivate managed AI agents directly from their records using kill switch protocols for immediate containment if necessary.

## Practical Use for ServiceNow Customers

ServiceNow customers can leverage the AI asset inventory to gain a full understanding of their AI landscape, maintain compliance, and enforce governance policies effectively. By tracking AI assets' lifecycle and risk levels, teams can prioritize oversight, automate management tasks, and ensure AI systems align with organizational standards and regulatory requirements. The inventory's filtering and role-based access features support focused governance activities, while evaluation metrics and bulk management capabilities simplify ongoing asset maintenance and risk mitigation.  
Track every AI system, model, prompt, and dataset across your enterprise and control which assets participate in governance, monitoring, and risk workflows.

## Key benefits {#disc-ai-asset-inventory__key-benefits}

The AI asset inventory is the central record of AI assets in your organization. Assets that AI Control Tower discovers or identifies through traces appears here, organized by asset type.

* Get a complete picture of every AI asset across your organization, including assets discovered automatically and those added manually or through connectors.
* Control which assets participate in governance, monitoring, and risk workflows by setting management status directly from the inventory list.
* Filter and refine the asset list by type, risk classification, lifecycle state, asset status, and management status to focus on what needs attention.
* Add assets manually when a connector or trace can't reach them. See [Creating AI assets manually](https://www.servicenow.com/docs/FieyViWbwNt6WEmeyPm32A "You can create AI assets manually to track and manage the life cycles of your AI systems, AI models, prompts, and datasets.").

{#disc-ai-asset-inventory__ul_key_benefits}

For the ways AI Control Tower adds assets to your inventory automatically, see [Discovering AI assets](https://www.servicenow.com/docs/17cMVq5y9APut3M4KNBsmg "Populate your AI asset inventory automatically as AI Control Tower discovers AI assets across external platforms, hyperscaler traces, and your ServiceNow instance.").
Figure 1. AI asset inventory

## AI assets list columns {#disc-ai-asset-inventory__ai-assets-list-columns}

{#disc-ai-asset-inventory__table_ux4_5dv_kkc__entry__2}

| Column | Description |
|-|-|
| Name | The name of the AI asset as registered in AI Control Tower. |
| Asset type | The category of the AI asset, such as AI dataset, AI prompt, AI model, MCP server, or AI system. |
| State | The operational state of an AI asset. |
| Status | The governance status of the asset within its current phase. |
| Managed status | Indicates whether the asset is actively governed. Displays as true or false. |
| Domain | Domain the asset belongs to. |
| Managed by | The team or individual responsible for governing the asset. |
| Risk classification | The assessed risk level of an AI asset. |
| Updated | The date and time the asset record was last modified. |
[ ]

{#disc-ai-asset-inventory__table_ux4_5dv_kkc}

## Required roles {#disc-ai-asset-inventory__required-roles}

The AI steward \[sn_ai_governance.ai_steward\] or AI system owner \[sn_ai_asset_mgmt.ai_asset_owner\] role is required to access the AI asset inventory.

## Accessing the AI asset inventory {#disc-ai-asset-inventory__accessing}

Navigate to AllAI Control TowerHomeInventoryAssets.

## AI systems {#disc-ai-asset-inventory__ai-systems}

An AI system is a deployed AI application or capability that your organization uses to perform work. Examples include Now Assist skills, AI agents built in AI Agent Studio, and AI-powered applications running on external platforms. AI systems are the primary unit of governance in AI Control Tower, and the assets you classify for risk, assign stewards to, monitor for performance, and measure for value.

In the Inventory, AI systems are identified by asset type: Agentic AI, Generative AI, or Classic AI. Select the AI systems tab to filter to these three types together.

* Filter to AI systems only to focus on the assets central to your governance program, regardless of provider.
* Select an AI system record to review lifecycle details, risk classification, monitoring data, and value metrics in one place.
* Designate AI systems as managed to bring them into governance workflows, or initiate a steward review to transition an unmanaged system automatically.
{#disc-ai-asset-inventory__ul_ai_systems}

## AI models {#disc-ai-asset-inventory__ai-models}

An AI model is a machine learning model or large language model (LLM) that powers one or more AI systems. Tracking models in the inventory gives you visibility into which models are in use across your organization, which vendors
supply them, and whether they are approved for their current deployment context.

* Identify which models are active across your organization, including models imported from external providers such as OpenAI, Mistral AI, and Google.
* Review provider, vendor, and lifecycle state for each model to confirm models are from approved sources and are current.
* Spot models that may need review as vendors release new versions or as vendor relationships change.
{#disc-ai-asset-inventory__ul_ai_models}

## AI datasets {#disc-ai-asset-inventory__datasets}

A dataset is a collection of data used for training, fine-tuning, or evaluating an AI model. Tracking datasets in the inventory supports data lineage, compliance documentation, and risk assessment for AI systems that depend on
specific training data.

* Track training, fine-tuning, and evaluation datasets alongside the models they support to maintain data lineage across your AI portfolio.
* Review provider, asset state, and management status for each dataset to identify datasets that may affect risk classification for dependent AI systems.
* Support compliance documentation by keeping a complete record of the data your AI models were built on.
{#disc-ai-asset-inventory__ul_datasets}

## AI prompts {#disc-ai-asset-inventory__prompts}

A prompt is a system prompt or prompt template that shapes the behavior of an AI system. Prompts define the instructions, context, and constraints an AI model operates under, making them a key governance asset for organizations
that require AI behavior aligns with policy and acceptable use standards.

* See all system prompts and templates in use across your organization's AI systems in one place.
* Identify prompts that may need policy review or update as acceptable use standards evolve.
* Understand the reach of any prompt change before making it by tracking which AI systems and models each prompt supports.
{#disc-ai-asset-inventory__ul_prompts}

## Reviewing duplicate AI assets {#disc-ai-asset-inventory__reviewing-duplicates}

AI asset deduplication can detect assets in your inventory that perform the same function, such as when
two teams independently build similar AI agents to solve the same problem. AI Control Tower surfaces these likely duplicates as a group for an AI steward to review. The AI steward can then confirm the group as true duplicates or dismiss it, before changes are made to relevant asset
records. For more information, see [Duplicate AI assets](https://www.servicenow.com/docs/~LourM1X4mVMisZE3r4e6g "AI asset deduplication in AI Control Tower identifies duplicate records in the AI asset inventory for AI Stewards to review and consolidate.").

## Managing the status of AI assets {#disc-ai-asset-inventory__section_esd_pml_hjc}

Control whether an AI asset participates in governance workflows by setting its management status. Managed assets are included in lifecycle management, governance, risk classification, value assessment, security, and privacy
capabilities. Unmanaged assets are tracked in the inventory but excluded from those workflows. For more information, see [Managed and unmanaged AI assets](https://www.servicenow.com/docs/4KhGkhnsOGMHlOIxk6m9VQ "Control whether an AI asset participates in governance workflows by changing the status in AI Control Tower.").

## Managing AI assets in bulk {#disc-ai-asset-inventory__section_bulk_managing_assets}

Classify large numbers of AI assets automatically instead of moving them to managed one at a time by creating automation rules. A rule matches assets against conditions you define, such as asset type or usage patterns, and marks
every matching asset as managed at each scheduled run. For more information, see [Managing AI assets in bulk](https://www.servicenow.com/docs/2eOpQu201LEdpK6R1_uZSg "Classify large numbers of AI assets automatically instead of moving them to managed one at a time by creating automation rules.").

## Managing evaluations {#disc-ai-asset-inventory__section_jn4_sll_hjc}

Detect quality and safety regressions in your managed AI assets by enabling evaluation. Evaluation scoring assesses AI interactions against configurable quality and safety metrics and surfaces trends so you can investigate
regressions before they affect users. Disable evaluation when an asset no longer needs active scoring; historical scores are retained either way. For more information, see [Evaluating AI systems](https://www.servicenow.com/docs/Dg1sdX3eoUUwrM22n8G69Q "Monitor the quality and safety of select AI systems by enabling evaluations.").

## Deactivating managed AI agents {#disc-ai-asset-inventory__section_deactivate}

Deactivate a managed AI agent directly from its asset record using AI agent containment with
kill switch protocol. For more information, see [Deactivate a managed AI agent](https://www.servicenow.com/docs/mIFBSZ3EWjDJXgL5y1jctA "Deactivate a managed AI agent directly from its asset record using kill switch protocol, even if the agent has no associated security event.").

*[\>]: and then