Enterprise network operations teams are being asked to manage increasingly complex environments with an expanding set of monitoring and observability tools. A survey of 356 enterprise IT professionals found that organizations typically run six to eight observability tools, a level of tool sprawl that creates integration headaches, staffing pressure, and gaps in end-to-end visibility as teams turn to artificial intelligence for help.
The research highlights a paradox facing IT leaders. AI can correlate alerts, summarize incidents, and detect anomalies across large volumes of telemetry, but it does not remove the underlying operational problems created by fragmented tooling and inconsistent data. In many cases, AI makes those problems more visible and more expensive to ignore.
Tool sprawl becomes an operational tax
Observability has evolved from simple infrastructure monitoring into a broad discipline that spans networks, applications, cloud platforms, databases, and user experience. Each domain often brings its own tools, data models, and workflows. The result is an environment in which engineers must move between dashboards, reconcile conflicting signals, and manually stitch together context during incidents.
The survey quantified the operational tax. Forty-five percent of respondents cited skills and staffing burdens from operating multiple tools. Forty-four percent said integration and API complexity consume engineering time. Forty-one percent said context switching among tools slows investigations and response. Forty percent reported increased manual effort, and 34 percent said alert noise creates toil and cognitive overload.
Those numbers point to a deeper problem than dashboard fatigue. Tool sprawl raises total cost of ownership, slows incident detection and resolution, creates inconsistent data quality, and leaves coverage gaps. It also undermines end-to-end visibility across technology domains. A network team may see healthy connectivity, while cloud and application teams see normal operations in their own tools, even as end users experience a serious problem. Every team can be correct within its silo while the business impact remains unresolved.
That scenario is common because teams often lack the ability to correlate data across tools and recognize an anomaly tied to the incident. The issue is not necessarily a lack of monitoring. It is a lack of shared context and coordinated analysis.
Unification efforts advance without a single pane of glass
Many enterprises are trying to unify observability, but few expect to consolidate everything into one platform. According to the survey, 51 percent of respondents had tool unification efforts under way, 32 percent were still planning or evaluating the approach, and 17 percent had completed unification. Larger organizations reported using more observability tools, which makes integration and governance more difficult.
The research also challenges the promise of a single pane of glass. One of the clearest findings is that no organization realistically achieves one interface that covers every domain, every team, and every data source. The goal instead is to connect tools, standardize data, and create workflows that let teams see across silos without pretending those silos no longer exist.
That distinction matters for network operations. Network telemetry, application performance data, cloud metrics, security events, and service management records often live in separate systems with different owners. Unification requires more than a new dashboard. It requires common data models, APIs, event correlation, and agreed-upon processes for incident response.
AI adoption grows alongside data fragmentation
Enterprises are already applying AI to observability data. Thirty-two percent of respondents said they use AI extensively across observability tools, 51 percent use it in selected areas, and 13 percent are still piloting it. Reported use cases include incident summaries and explanations, alert correlation and noise reduction, capacity forecasting, anomaly detection, predictive analytics, and root-cause analytics.
AI is particularly attractive for alert correlation and anomaly detection because modern environments generate more signals than humans can reasonably process. AI can group related alerts, identify patterns, and reduce the time spent on routine triage. It can also produce incident summaries that help responders understand what happened without reading every dashboard.
But AI depends on good data. Fragmented or inconsistent data limits model effectiveness, and organizations also cited concerns about inaccurate outputs, security risks, and compliance exposure. AI can help overcome data fragmentation across tools, but it cannot overcome bad data and bad operations in general. If the underlying telemetry is incomplete, inconsistent, or poorly governed, AI may simply produce faster but less reliable conclusions.
The survey found that organizations using AI extensively also reported greater confidence in their observability unification strategies. That correlation does not prove AI causes better unification, but it suggests that mature data practices and AI adoption tend to reinforce each other. Organizations with cleaner data, stronger integration, and clearer ownership are better positioned to get value from AI.
Governance separates leaders from laggards
Tool governance is emerging as a critical prerequisite for simplification. The problems associated with tool sprawl can compound one another: integration complexity consumes engineering time, staffing burdens slow projects, context switching delays response, and alert noise adds toil. Strong governance gives organizations a way to prioritize tools, define ownership, and reduce duplication before they attempt a broader unification program.
About one-quarter of organizations, 24 percent, reported fully centralized observability tool governance, while 48 percent said ownership was centralized for the most part with some domain-specific exceptions. Those with centralized ownership reported greater confidence in their unification strategies, deeper integration between observability and IT service management, more effective incident-impact analysis, and more mature use of AI-powered observability.
Centralization, however, is not just an org-chart change. Understaffed tool governance and engineering teams had less success with unification strategies. Putting one group in charge without giving it resources, authority, and executive support can create a bottleneck rather than a center of excellence. Centralized authority must come with the capacity to implement and deliver.
Agentic AI raises the stakes for control
AI governance becomes more important as AI systems move from analysis to action. When AI only summarizes alerts or recommends a root cause, a bad output can waste time. When AI can change firewall rules, modify configurations, or remediate infrastructure, a bad output can disrupt business services.
Agentic AI systems are probabilistic, meaning their responses and actions are not always predictable. Many traditional cybersecurity controls are deterministic, following predefined rules. That mismatch creates a governance challenge. An AI system might identify a firewall rule as unnecessary, for example, without understanding that the rule supports a legitimate business process that runs for one day each quarter. The rule may look like a mistake to a model trained to eliminate redundant access, but removing it could break a critical workflow.
Scenarios like that show why human review remains necessary for unusual exceptions, business-critical rules, and potentially harmful or inaccurate outputs. Controls should be tailored to an organization’s risk profile, with clear limits on what AI systems can change, when they can act, and what approvals are required. Network operations teams will need to integrate observability data and workflows while maintaining strong tool governance and explicit controls over AI behavior.
The survey findings point to a practical agenda: reduce unnecessary tool overlap, standardize data and integrations, centralize ownership with adequate resources, and define where AI can act autonomously and where humans must remain in the loop. AI is another tool in the operations tool chest, and it will continue to evolve. The organizations that gain the most will be those that govern it as carefully as they govern the infrastructure it is allowed to change.
Source: Network World News