Autonomous Data Hygiene: The Self-Cleaning CMDB

The biggest "hassle" in CMDB management has always been the manual effort required to keep records accurate. Zurich introduces Now Assist for CMDB, featuring AI agents that perform continuous data sanitation.

Automated Duplicate Management: Using semantic reasoning, AI agents can identify duplicate Configuration Items (CIs) that traditional discovery might miss. For example, if a cloud instance and an on-premise server record share identical serial numbers and MAC addresses, the AI doesn't just flag them—it executes a merge strategy that preserves the most accurate attributes from the highest-trust source.

Intelligent Lifecycle Tracking: AI agents now monitor "signals" such as discovery last-seen dates and incident activity. If a server has been "dark" for 90 days with no related change requests, the AI autonomously proposes a "Retired" status, preventing stale data from cluttering your environment.

Agentic Discovery: Beyond Static Scans"

Traditional discovery is often slow and restricted by schedules. In the Zurich release, Agentic Discovery works in real-time through the AI Agent Fabric.Instead of waiting for a weekly scan, "Build Agents" and "Discovery Agents" use Zero Copy connectors to tap into cloud APIs (AWS, Azure, GCP) the millisecond a new resource is provisioned. This ensures that your CMDB is a living reflection of your infrastructure. Furthermore, Zurich introduces Credential-Free Discovery for specialised environments like Kubernetes, allowing agents to map deep patterns and microservice dependencies without the security risk of managing thousands of individual service accounts.

CMDB Summarisation and Predictive Insights

Understanding complex CI relationships used to require a degree in data science. Now, it just requires a conversation.

AI CI Summarisation: With a single click, an admin can generate a human-readable summary of a CI. Instead of looking at a table of raw data, you get a concise paragraph: "This is a production SQL database hosted on AWS, currently supporting the Global Payroll Service, with three critical security vulnerabilities and a scheduled maintenance window in two days.
Predictive Data Quality: Using historical trends, AI agents can predict which CIs are likely to become "stale" or which relationships are missing. By analyzing traffic flow logs, the AI can suggest new relationships between components that were never manually mapped, ensuring your Service Mapping is always complete.

The Foundation for Trust: AI Control Tower

In 2026, the CMDB is the primary training ground for your AI agents. If the CMDB is wrong, your AI’s decisions will be wrong. The AI Control Tower in Zurich acts as the guardian of this data.It monitors the "health" of the data used by your AI agents, ensuring they aren't hallucinating resolutions based on outdated CI info. By maintaining high Data Quality Scores through automated audits, the Control Tower ensures that your ITSM and ITOM agents are acting on "Authoritative Truth," reducing the risk of automated errors in production environments.

Data That Works for You

The Zurich release confirms that a manual CMDB is a failed CMDB. By automating the "grunt work" of de-duplication, mapping, and lifecycle management, ServiceNow has made the CMDB the most valuable asset in the IT stack. It is no longer a database you have to maintain; it is an intelligent partner that maintains itself.