Enterprises are deploying AI agents faster than they can govern them. At Knowledge 2026 last week, ServiceNow announced a complete redesign of its AI Control Tower—transforming it from a simple monitoring dashboard into the first unified governance layer that works across AWS, Azure, Google Cloud, SAP, Oracle, Workday, and 25+ other enterprise systems.

The problem they're solving is urgent. CIOs are managing hundreds of AI agents across disconnected platforms with no unified view of what these agents are doing, who authorized them, what data they're accessing, or how much they're costing. ServiceNow's bet is that if you can't see it, you can't govern it—and if you can't govern it, AI becomes a compliance nightmare instead of a business advantage.

The "Runaway AI Spend" Problem

ServiceNow CEO Bill McDermott framed the challenge bluntly: "We are the rules and rails of business." The company is positioning AI Control Tower as the governance backbone that lets enterprises deploy AI at speed without losing financial or security control.

Here's what "runaway AI spend" looks like in practice: A Fortune 500 company deploys 50 different AI agents across marketing, finance, IT, and HR. Each team picks their own vendor—some use Microsoft Copilot, others use Claude, a few build custom agents on AWS Bedrock. Six months later, the CFO discovers they're spending $2.3 million annually on model API calls with zero visibility into which agents are driving ROI and which are burning budget on low-value tasks.

The new AI Control Tower addresses this across five operational dimensions:

1. Discover - Find every AI agent, model, and identity across the enterprise (not just ServiceNow deployments)
2. Observe - Monitor runtime performance, token usage, and API calls in real-time
3. Govern - Enforce policies, approvals, and compliance rules at the point of execution
4. Secure - Verify identity, enforce least-privilege access, and maintain audit trails
5. Measure - Track cost per agent, ROI dashboards, and business value attribution

This isn't just for ServiceNow's own AI agents. The platform now integrates with Anthropic's Claude Cowork, Microsoft's Agent 365 ecosystem (including Copilot Studio and Azure Foundry agents), and NVIDIA's OpenShell runtime for homegrown enterprise agents.

How It Works: Action Fabric and MCP Server

The technical breakthrough is Action Fabric. ServiceNow created a new execution layer that lets any AI agent—whether built on Claude, Microsoft Copilot, or a custom internal tool—trigger governed ServiceNow workflows without requiring users to open a ticket or manually execute tasks.

Jon Sigler, EVP and GM of ServiceNow's AI Platform, explained the architecture: "Enterprises are under pressure to deploy AI and show results, but there's a major gap between adoption and accountability. AI Control Tower provides unified governance across the entire enterprise AI stack, so security and control move at the speed of the business."

Here's a concrete example: An AI agent built on Claude detects a security vulnerability in production. Instead of sending an alert that requires a human to log into ServiceNow, open a ticket, assign it to the right team, and kick off a remediation workflow, the Claude agent directly triggers ServiceNow's automated incident response playbook via Action Fabric. AI Control Tower validates the agent's identity, checks that it has permission to trigger that specific workflow, logs the action for audit purposes, and tracks the cost of the model inference that initiated the action.

This is what Boris Cherny, Head of Claude Code at Anthropic, meant when he said: "The gap between knowing what needs to happen and making it happen is where productivity dies. Connecting Claude Cowork to ServiceNow's system of action closes that gap by enabling enterprise execution, directly in the flow of work."

The governance layer operates through the Model Context Protocol (MCP) Server, which ServiceNow uses to expose its 100 billion workflows and 7 trillion annual workflow transactions to external AI agents while maintaining identity verification, permission control, and full auditability.

The Multi-Vendor Reality

ServiceNow isn't trying to replace your existing AI vendors. They're positioning themselves as the orchestration and governance layer that sits on top of everyone else's AI infrastructure.

The redesigned AI Control Tower now integrates with:

• Microsoft Agent 365 - Governs Copilot Studio and Azure Foundry agents, with ServiceNow AI specialists appearing in Microsoft's Agent 365 Marketplace
• Anthropic Claude - Claude Cowork agents can execute ServiceNow workflows directly via Action Fabric
• NVIDIA OpenShell - Project Arc autonomous desktop agents run in NVIDIA's secure runtime while being governed by AI Control Tower
• AWS, Azure, Google Cloud - Multi-cloud model deployments monitored and governed through a unified interface
• SAP, Oracle, Workday - Cross-application governance for enterprise SaaS AI deployments

Charles Lamanna, EVP for Copilot, Agents, and Platform at Microsoft, confirmed the partnership: "Together, we're helping customers act on insights more quickly and drive meaningful outcomes across their business processes."

For CIOs evaluating AI governance options, this multi-vendor strategy matters. If you've already standardized on Microsoft for productivity, SAP for ERP, and Workday for HR, you don't want to rip-and-replace those systems just to get AI governance. ServiceNow is betting that enterprises will pay for a unified governance layer that works across all of them.

Cost Tracking: The CFO Perspective

Finance leaders care about one question: What are we spending, and what are we getting?

AI Control Tower's new cost tracking and ROI dashboards are designed to answer that question at the agent level. CFOs can now see:

• Total AI spend by department - Which teams are burning the most budget on model API calls
• Cost per agent - Individual agents ranked by monthly spend (identify the expensive ones)
• ROI attribution - Business outcomes linked to specific AI deployments (revenue generated, hours saved, tickets closed)
• Spend trends - Month-over-month growth in AI costs, with alerts when spending accelerates beyond policy thresholds

ServiceNow set a $30 billion subscription revenue target for 2030, with AI solutions expected to represent more than 30% of annual contract value. That projection assumes enterprises will pay for governance as AI sprawl becomes unmanageable—and the company is betting that multi-vendor governance is the wedge that justifies ongoing subscription spend.

HDFC Bank, India's largest private sector bank, is already using AI Control Tower in production. Ramesh Lakshminarayanan, Group CIO, explained the governance imperative: "As India's largest private sector bank, AI governance is foundational. AI Control Tower is the common governance layer across all of it, giving us the visibility to manage every AI use case and the confidence to scale."

Security and Risk: The Armis + Veza Integration

ServiceNow's $1 billion cybersecurity business (reached in 2025) is now fully integrated into the AI governance strategy. The company consolidated its two major security acquisitions—Armis (asset intelligence) and Veza (identity governance)—into a new Autonomous Security and Risk offering.

Here's how the integration works:

• Armis provides agentless, real-time asset intelligence across IT, OT, IoT, code, and connected devices, feeding directly into ServiceNow's Configuration Management Database (CMDB)
• Veza maps every identity (human and non-human AI agents), enforcing least-privilege access at the point of action and providing full audit trails

John Aisien, SVP and GM for Security and Risk at ServiceNow, framed the value proposition: "Today's CISOs have to not only neutralize threats in real time but also report risk to the board with conviction. Autonomous Security & Risk replaces that fragmented stack with a single graph that maps every identity, every permission, and every connected asset."

Real-world impact numbers ServiceNow shared:

• A global energy company (70+ countries) cut threat containment time by 97%
• A major U.S. financial institution eliminated 96% of dormant non-human identities
• A Fortune 100 aerospace manufacturer reduced control attestation time by 75%

For regulated industries—financial services, healthcare, government—this identity governance layer addresses the "who authorized this AI agent to access customer data?" question that auditors and compliance teams are now asking every quarter.

Production Deployments: What's Actually Working

ServiceNow isn't just selling vaporware. The company shared several production deployments to back up its governance claims:

• Internal L1 IT Service Desk AI Specialist handles cases 99% faster than human agents
• Autonomous CRM manages over 100 million customer cases monthly, coordinates 16+ million orders, and configures 7+ million quotes per month
• PayPal reports that "database tasks are twice as fast, and our longest-running operations are five times faster" after deploying ServiceNow's AI Platform

Amit Zavery, President and Chief Product Officer at ServiceNow, emphasized the production-first approach: "Enterprises need AI that senses, decides, and acts securely in line with organizational guardrails. With ServiceNow expanding Autonomous Workforce across critical business functions, organizations can deploy AI specialists to act at scale from a single, governed platform."

The L1 Service Desk example is particularly telling. ServiceNow's internal IT team uses AI specialists to handle 99% of tier-1 support tickets autonomously, with AI Control Tower governing every action the agent takes—including ticket assignment, knowledge base searches, automated resolutions, and escalation workflows. This is the same governance framework they're now selling to enterprise customers.

What CIOs Should Do Next

If you're managing more than 10 AI agents across your organization, here's the action plan:

1. Audit your AI sprawl - How many AI agents are currently deployed? Which teams own them? What systems do they access? What's the total monthly cost?

2. Identify your governance gaps - Can you answer these questions: Who authorized each agent? What data can it access? What actions can it take? What's the audit trail if something goes wrong?

3. Evaluate multi-vendor governance - If you're using Microsoft, AWS, Google Cloud, SAP, and other platforms, does your current governance approach work across all of them, or are you managing each silo separately?

4. Calculate your AI ROI - Do you have cost-per-agent data? Can you tie AI spend to business outcomes? Can you identify which agents are driving value and which are burning budget?

5. Test ServiceNow's Innovation Lab (available May 2026, general availability August 2026) - If you're already a ServiceNow customer, request access to the expanded AI Control Tower and run a 90-day pilot with your highest-spend AI agents.

The NVIDIA partnership adds another layer worth watching. Kari Briski, VP of Generative AI for Enterprise at NVIDIA, announced Project Arc—enterprise autonomous desktop agents protected by NVIDIA's OpenShell runtime and managed by AI Control Tower. If NVIDIA's secure agent runtime becomes the industry standard for on-premises AI deployments, ServiceNow's governance integration becomes even more valuable for hybrid cloud enterprises.

The Bottom Line for Business Leaders

AI governance isn't a nice-to-have anymore—it's table stakes for CFOs and CISOs. As AI agent deployments scale from dozens to hundreds across enterprise functions, the cost and risk exposure grows exponentially without unified governance.

ServiceNow's AI Control Tower is the first serious attempt at multi-vendor AI governance that works across cloud providers, enterprise SaaS platforms, and custom-built agents. The question for enterprise leaders is whether ServiceNow's 20-year operational data advantage (100 billion workflows, 7 trillion transactions per year, mature CMDB integration) makes them the right governance layer—or whether competitors like Microsoft, AWS, or Google Cloud will build equally capable governance into their own platforms.

For now, ServiceNow has first-mover advantage in cross-platform AI governance. If you're spending more than $500K annually on AI infrastructure and you can't answer basic questions about agent costs, permissions, and ROI, that's a governance gap worth addressing this quarter.

---

Continue Reading

• AI Agent Identity Crisis: 92% of CISOs Can't Track Non-Human Accounts
• The $176M Problem: Why Humans Reject AI Recommendations
• SAP's €100M AI Push: 200 Agents Automate Finance Operations

---

About THE D*AI*LY BRIEF: Twice-weekly analysis of Enterprise AI for technical and business leaders. Written by Rajesh Beri, Head of AI Engineering.

Connect:

• LinkedIn: linkedin.com/in/rberi
• Twitter/X: @rajeshberi
• Facebook: facebook.com/rajeshberi