When ServiceNow introduced AI Control Tower at Knowledge 2025, the company was responding to a specific enterprise problem. AI was spreading across the organization faster than governance could keep up.
In this context, AI Control Tower was positioned as a centralized command center to govern, manage, secure, and realize value from AI agents, models, and workflows.
However, only a year later, the context changed. At Knowledge 2026, ServiceNow expanded its framing around what it called “AI chaos.” Enterprises were moving toward environments with hundreds of AI applications, agents, and models operating across different systems and vendors. So ServiceNow’s focus shifted toward governing, securing, and measuring AI across the enterprise.
The rise of agentic AI helps explain why that shift matters. Let’s consider a recent experiment with OpenClaw to summarize the issue.
Wired journalist Will Knight connected the agent to email, messaging platforms, a browser, and shopping tools. At one point, the agent kept trying to buy a single serving of guacamole after being told to continue with the rest of the grocery list.
It is a funny story when the stakes are one person’s grocery order. In an enterprise environment, an agent with access to business systems, sensitive data, and operational tools could create a very different kind of incident.
AI is becoming part of enterprise environments, yet its inventory and governance often remain distributed across individual platforms. One team may be running AI through ServiceNow. Another may be using AWS, Azure, or Google Cloud. Others may rely on Salesforce, Snowflake, internal applications, or agentic frameworks.
Each platform can provide its own controls, but the enterprise still needs a way to understand the bigger picture:
- What AI exists.
- Who owns it.
- What systems it can access.
- What risks it carries.
- What value it delivers.
In short, as AI becomes more deeply embedded in how companies operate, visibility becomes an increasingly important part of running the business. That is the environment in which AI Control Tower enters the market.
Let’s examine it’s governance framework, technical capabilities, licensing model, relationship with AI Gateway, and approach to enterprise implementation.
Why ServiceNow AI Control Tower?
Now, ServiceNow has long been used to manage applications, services, assets, workflows, risks, and operational processes. AI Control Tower extends that environment to the AI systems, becoming part of those same operations.
That gives us a different starting point for AI governance.
An AI model or agent can be connected to the business services, applications, assets, owners, and processes around it. Instead of maintaining an isolated AI inventory, organizations can place their AI estate within the broader context of how the business operates.
This is where the CMDB becomes particularly important. AI assets can be associated with existing enterprise services and assets, giving teams a way to understand where an AI system fits within the organization and what could be affected if it creates a risk or fails.
The approach also reflects how AI Control Tower has evolved since its release in 2025.
ServiceNow initially positioned the product around centralized governance for AI agents, models, and workflows. Today, the company expanded that vision around discovering, observing, governing, securing, and measuring AI across the enterprise.
That evolution matters because enterprise AI is becoming increasingly distributed.
A company may use ServiceNow AI alongside models and agents running across third-party platforms or internally developed environments. AI Control Tower is designed to bring information from these different environments into a common governance framework.
This way, the business value comes from connecting several disciplines:
- AI governance: Organizations can establish ownership, approvals, policies, and lifecycle processes for AI assets across the enterprise.
- Risk and compliance: AI initiatives can be assessed against organizational controls and regulatory requirements as part of existing governance processes.
- Security and privacy: AI assets can be evaluated in the context of identities, access, sensitive data, and the systems they interact with.
- Business strategy: AI initiatives can be connected to strategic priorities, helping leaders evaluate where AI investments belong in the broader portfolio.
- Value management: Organizations can track adoption, productivity, costs, and business outcomes alongside the AI assets generating them.
AI Control Tower vs. AI Gateway: Where Do They Fit?
While they are part of the same broader architecture in ServiceNow, they operate at different levels. The first one properly refers to a broader governance framework. The second one is one of the capabilities within it.
This distinction is important because the ServiceNow AI gateway addresses a specific technical problem within the larger AI governance model. But let’s check the technical difference between them.
As we saw, AI Control Tower helps teams discover AI assets, manage their lifecycle, assess risk, establish governance policies, and measure business value. An AI gateway, on the other hand, provides a more focused control point for interactions between AI agents and external tools through the Model Context Protocol (MCP).
In practical terms, it sits between an AI agent and the external tools that agent is allowed to access. It can apply authentication, access controls, monitoring, and security policies to those interactions. Within ServiceNow, these capabilities become part of the broader governance environment provided by AI Control Tower.
This distinction also helps clarify LLM governance. The ServiceNow AI Gateway is not the general-purpose mechanism to determine which provider should power every AI capability. That function is handled through AI Model Providers within the ServiceNow AI environment.
ServiceNow allows organizations to configure approved model providers for its AI capabilities. Administrators can control which providers are available and assign them at different levels, including instances, skill groups, and individual skills.
This creates two related control mechanisms within the broader governance architecture:
- AI model providers: Govern which models and providers ServiceNow AI capabilities are allowed to use.
- AI gateway: Governs how AI agents interact with external tools and services through MCP.
The distinction becomes particularly relevant as enterprises adopt more agents and connect them to a growing number of external systems.
In short, AI Model Providers address the model layer. AI gateway addresses the agent-to-tool interaction layer. AI Control Tower provides the broader governance framework that brings these capabilities together with risk, security, strategy, and value management.
How Is AI Control Tower Licensed?
The licensing model of AI Control Tower is also based on the AI assets being managed, rather than on the number of users or the number of prompts generated. ServiceNow uses Subscription Units, with one managed AI model, AI system, or dataset corresponding to one Subscription Unit.
This way, if the number of assets under governance grows, the required subscription capacity can grow with it.
This set AI Control Tower apart from the consumption models used for many AI services. Subscription Units measure the AI estate being governed; they do not measure LLM token consumption. However, Now Assist has its own licensing and consumption model, so the two should be evaluated separately.
On the other hand, this should also be viewed as a set of capabilities and dependencies. The core Control Tower environment can be extended with capabilities for areas such as AI risk, compliance, security, and asset management, depending on the organization’s requirements and entitlements.
For an enterprise evaluating the product, the practical calculation therefore starts with three questions:
- Which AI assets need to be governed?
- Which Control Tower capabilities are required?
- Which underlying AI products have their own licenses or consumption costs?
The answers determine the actual licensing footprint and help separate the cost of governing AI from the cost of using AI.
Building Your AI Control Tower Strategy
AI governance touches architecture, security, risk, data, workflows, and business strategy. Implementing AI Control Tower therefore requires the right combination of ServiceNow expertise and enterprise context.
As official ServiceNow partners, at Inclusion Cloud we can help you assess your AI governance requirements, configure and integrate AI Control Tower, connect it with your existing ServiceNow environment, and build the workflows needed to operationalize governance.
Our certified architects, implementation specialists, developers, administrators, and project managers can support the full delivery lifecycle, from advisory and implementation to continuous innovation. We can also provide dedicated ServiceNow talent or agile delivery pods through our inMOVE™ model.
Book a discovery call and let’s see how to help your organization move from AI visibility to AI governance that works within the way your business already operates.
