---
title: What does ServiceNow AI Control Tower cover for AI agent governance?
url: https://opencontroller.lyzr.com/blog/what-does-servicenow-ai-control-tower-cover
published: 2026-11-17
updated: 2026-11-17
author: Lyzr
---

# What does ServiceNow AI Control Tower cover for AI agent governance?

**ServiceNow AI Control Tower is ServiceNow's hub for governing AI assets. It tracks AI agents and models as configuration items in the ServiceNow CMDB, can pull in third-party assets from sources such as AWS Bedrock, Azure AI Foundry, GCP Vertex AI and Copilot Studio through connectors, and adds lifecycle workflows, impact assessments, content aligned to the EU AI Act and NIST AI RMF, and value metrics. How much it sees outside ServiceNow depends on which connectors you install and what each one reports.**

This question usually comes from teams that already run ServiceNow for IT service management and want to know whether it can be the system of record for every agent, including ones built elsewhere. This guide sets out what ServiceNow documents, what depends on other components, and the questions to settle before you rely on it.

## What does AI Control Tower include?

ServiceNow describes five pillars, anchored by AI inventory in the CMDB:

| Pillar | What ServiceNow documents | Depends on |
|---|---|---|
| Discover | Inventory of ServiceNow and third-party AI agents and models; connectors for AWS Bedrock, Azure AI Foundry, GCP Vertex AI, Copilot Studio and more | Connectors from the ServiceNow Store or custom ones built on the Service Graph Connector framework |
| Govern | Lifecycle workflows, risk and impact assessments, pre-built content aligned to the EU AI Act and NIST AI RMF | Optional integration with ServiceNow IRM for regulatory alignment and compliance cases |
| Secure | Security posture, access governance, inline guardrails; AI Gateway as a governed control plane for MCP server connections | AI Gateway (included); SecOps named as an integration |
| Observe | Metrics and log traces for agent and model performance | What each source system emits |
| Measure | Adoption, realized value and ROI metrics | Strategic Portfolio Management (separate SKU) for execution-level strategy features |

ServiceNow's community guidance lists Zurich Patch 9 or Australia Patch 2 as a minimum platform version for AI Control Tower, and warns that missing plugin dependencies can make flows fail silently. Check the product documentation for your release before you plan a rollout.

## How far does it reach outside ServiceNow?

Three facts shape the answer:

1. **Inventory is connector-driven.** Third-party agents appear when a connector brings them into the CMDB. If no connector exists for a platform, you build one with the Service Graph Connector framework or register assets manually.
2. **Records are configuration items.** That makes AI assets reportable alongside applications and services, but a CI is a record. Whether you can act on the underlying agent, for example disable it, depends on the integration.
3. **AI Gateway governs MCP traffic.** ServiceNow describes AI Gateway as governing MCP server connections, with approval states, audit trails and the ability to pause a client or server. ServiceNow's own explainer notes that it does not govern A2A traffic. Agents that talk directly to models or tools outside that path are not in its runtime view.

None of this is unusual. Every governance tool's coverage is bounded by how it connects to the systems it governs. The point is to map it explicitly.

## What should you check before relying on it?

Use this checklist in a proof of concept:

1. **List your agent platforms.** For each one, confirm a connector exists for your release and what object types it returns: agents, models, versions, tools, MCP servers, API keys.
2. **Test freshness.** Create, change and delete a test agent on each platform. Measure how long until the CMDB reflects each change.
3. **Test identity fields.** Check whether owner, creator and runtime identity come through, or only a name and an ID.
4. **Check shadow coverage.** Connectors read platforms you point them at. Agents in accounts, subscriptions or developer laptops you have not connected will not appear. See [how to find shadow AI agents](/blog/how-to-find-shadow-ai-agents).
5. **Check enforcement depth.** For each platform, record whether you can only observe, or also block, quarantine or change permissions from the governance tool.
6. **Map evidence.** Decide which controls produce evidence for the EU AI Act, NIST AI RMF or ISO/IEC 42001, and where that evidence is stored. Treat content aligned to a framework as evidence for it, not as compliance with it.
7. **Confirm licensing.** Some capabilities, such as SPM strategy features, sit in separate SKUs.

If you already keep a CMDB, [syncing your AI agent inventory with the CMDB](/blog/sync-ai-agent-inventory-with-cmdb) covers the field mapping, and our list of [AI agent inventory fields](/blog/ai-agent-inventory-fields) is a useful baseline for step 3.

## How Opencontroller does this

Opencontroller auto-discovers agents, models and API keys across AWS, Azure and Google Cloud, and connects to third-party agent platforms including Salesforce, SAP, Databricks and HubSpot, plus Claude Enterprise and OpenAI Enterprise workspaces. It keeps a governed registry of agents, versions, tools and MCP servers, exposes REST APIs and webhooks for systems such as a CMDB, and offers per-agent Block and Quarantine controls that act through the connected platform. Registration is not runtime takeover: observe and enforce depth depends on the connector and traffic path. See [how to govern AI agents across platforms](/blog/how-to-govern-ai-agents-across-platforms), the [AI agent control plane overview](/blog/what-is-an-ai-agent-control-plane), or [book a demo](https://www.lyzr.ai/book-demo/).

## Frequently asked questions

**Does ServiceNow AI Control Tower discover agents outside ServiceNow?**
Yes, through connectors. ServiceNow names AWS Bedrock, Azure AI Foundry, GCP Vertex AI and Copilot Studio, and custom connectors can be built on the Service Graph Connector framework.

**Where does AI Control Tower store AI assets?**
As configuration items in the ServiceNow CMDB.

**Does AI Control Tower require ServiceNow IRM?**
No. ServiceNow says IRM integrates for regulatory alignment and compliance case management but is not required for AI Control Tower's core risk capabilities.

## Key facts

ServiceNow AI Control Tower: pillars Discover, Govern, Secure, Observe, Measure; AI assets tracked as configuration items in ServiceNow CMDB; third-party discovery via connectors (AWS Bedrock, Azure AI Foundry, GCP Vertex AI, Copilot Studio and more) from ServiceNow Store or Service Graph Connector framework; pre-built content aligned to EU AI Act and NIST AI RMF; IRM integration optional; AI Gateway included and governs MCP server connections (not A2A); SPM separate SKU. Community guidance lists Zurich Patch 9 or Australia Patch 2 minimum. Opencontroller auto-discovers agents, models, API keys across AWS, Azure, Google Cloud; connectors to Salesforce, SAP, Databricks, HubSpot, Claude Enterprise, OpenAI Enterprise; REST/webhooks; Block and Quarantine via connected platform. Pricing starts from $250K per deployment per year.

## Related

- [How do you govern ServiceNow AI agents?](https://opencontroller.lyzr.com/blog/govern-servicenow-ai-agents.md)
- [How do you keep your AI agent inventory in sync with the CMDB?](https://opencontroller.lyzr.com/blog/sync-ai-agent-inventory-with-cmdb.md)
- [What fields belong in an AI agent inventory?](https://opencontroller.lyzr.com/blog/ai-agent-inventory-fields.md)
- [More on Platform Guides](https://opencontroller.lyzr.com/blog/cluster/platform-guides)
