Key facts at a glance

• Enterprise organizations typically use six to eight observability tools, creating integration, staffing, and visibility challenges as network operations teams adopt AI to manage complex IT environments.
• A survey of 356 enterprise IT professionals found that 51% had tool-unification efforts underway, 32% were planning or evaluating unification, and 17% had completed it.
• Tool sprawl affects IT teams through skills and staffing burdens, integration and API complexity, context switching, manual effort, and alert noise.
• AI is already used in observability workflows: 32% of respondents use it extensively, 51% use it in selected areas, and 13% are still piloting it.
• Centralized tool governance is linked to greater confidence in unification strategies and more mature use of AI, but centralization must be backed by resources.
• As AI systems begin acting on infrastructure, organizations need human oversight for unusual exceptions, business-critical rules, and potentially harmful or inaccurate outputs.

Enterprise network operations teams are being asked to manage increasingly complex environments with a growing set of monitoring and observability tools. A recent survey of 356 enterprise IT professionals found that most organizations run six to eight observability tools, a level of tool sprawl that consumes engineering time, increases staffing demands, and slows incident resolution. The same research shows that AI is already entering observability workflows, particularly for alert correlation and anomaly detection, but it is not a shortcut around operational discipline. Instead, AI is exposing the need for better integration, cleaner data, stronger governance, and clearer human oversight.

Tool sprawl strains IT operations

The problem is not simply that tools are numerous. Each tool brings its own data model, API, dashboard, alerting logic, and administrative overhead. Engineers must learn multiple interfaces and translate information across systems. During incidents, that translation work becomes especially costly because teams need to correlate signals quickly. When one tool shows healthy connectivity, another shows normal cloud performance, and a third shows that applications are operating within expected thresholds, end users can still be experiencing a serious problem. The tools may all report that everything is fine, but the organization lacks the ability to connect the dots.

The survey identified several ways tool sprawl affects IT teams. Forty-five percent 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. Thirty-four percent said alert noise creates toil and cognitive overload. These are not isolated complaints. They reinforce one another: more tools create more alerts, more alerts create more context switching, and more context switching increases the time needed to resolve incidents.

Tool sprawl also contributes to higher total cost of ownership, slow incident detection and resolution, inconsistent data quality, coverage gaps, and poor end-to-end visibility across technology domains. The cost is not only licensing and infrastructure. It is also the human cost of maintaining integrations, training staff, reconciling conflicting data, and managing workflows that cross multiple platforms. In many organizations, the observability stack has grown organically as teams adopted specialized tools for network, cloud, application, security, and service management needs. Each addition may have solved a local problem while adding to the overall complexity of the operating environment.

Unification remains incomplete

Many enterprises are still working to unify their observability environments. The survey found that 51% of respondents had tool-unification efforts under way, while 32% were still planning or evaluating that approach and 17% had completed unification. Larger organizations reported that they had more observability tools in use across their environments. That pattern is unsurprising. Larger enterprises tend to have more technology domains, more specialized teams, more regulatory requirements, and more legacy systems. Each domain may have legitimate reasons for using a particular tool, but the cumulative effect can be a fragmented observability estate.

Unification does not necessarily mean reducing the environment to a single platform. In practice, many organizations will continue to use multiple tools because different teams and use cases require different capabilities. The goal is to integrate data and workflows so that teams can investigate incidents across domains without manually stitching together information from separate consoles. A single pane of glass is an appealing idea, but it is rarely achieved in complex enterprises. The more realistic objective is a coordinated set of tools governed by common standards, shared data models, and integrated workflows.

That distinction matters for network operations teams. Network data often sits alongside cloud telemetry, application performance data, security events, and service management records. Without integration, teams may have visibility into their own domain but limited visibility into the end-to-end service. The result can be finger-pointing between teams, delayed incident response, and a poor experience for end users. Tool unification efforts are meant to address that gap, but they require sustained engineering work and clear ownership.

AI takes on more of the workload

Enterprises are now applying AI to operational data. Thirty-two percent of respondents said they use AI extensively across observability tools, 51% use it in selected areas, and 13% are still piloting it. Use cases include incident summaries and explanations, alert correlation and noise reduction, capacity forecasting, anomaly detection, predictive analytics, and root-cause analytics. These capabilities can be valuable in environments where engineers are overwhelmed by data and alerts. AI can help identify patterns, group related signals, and surface likely causes faster than manual review.

Organizations that reported extensively using AI also said they had greater confidence in their observability unification strategies. That finding suggests a virtuous cycle: better integration and governance produce better data, and better data makes AI more effective. Fragmented or inconsistent data can limit AI effectiveness, and organizations also cited concerns about inaccurate outputs and security and compliance risks. AI can help overcome data fragmentation across tools, but it cannot overcome bad data and bad operations in general. If the underlying processes are inconsistent, the tooling is poorly integrated, or the data is incomplete, AI may simply produce faster but less reliable conclusions.

This is an important caution for network operations leaders. AI is not a replacement for operational discipline. It is an amplifier of the environment in which it operates. In a well-governed observability environment, AI can reduce alert noise, improve incident summaries, and help engineers prioritize work. In a poorly governed environment, AI can add another layer of complexity, create new dependencies, and introduce risks around accuracy, security, and compliance. The survey results point to a practical conclusion: AI adoption should be paired with continued investment in data quality, integration, and tool governance.

Governance becomes central

Researchers said the problems associated with tool sprawl can compound one another, making strong tool governance essential before organizations try to simplify their observability environments. Governance is not just an administrative function. It determines which tools are approved, how data is shared, who owns integrations, how alerts are managed, and how changes are evaluated. Without governance, tool sprawl tends to accelerate because teams adopt new tools to solve immediate problems without considering the long-term operational burden.

About one-quarter of organizations, or 24%, reported fully centralized observability tool governance, and 48% 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. The pattern is clear: centralized governance helps create consistency, but it must be designed carefully so that domain teams still have the flexibility they need.

Centralization also means more than putting one group in charge. Understaffed tool governance and engineering teams had less success with their unification strategies. It is critical not just to centralize authority but to give that centralized group the resources it needs to implement and deliver. Governance without capacity becomes a bottleneck. The central team must be able to define standards, review exceptions, maintain integrations, and support domain teams without becoming an obstacle to progress.

When AI starts acting on infrastructure

AI governance becomes more important when AI systems begin taking actions across network infrastructure. As enterprises give AI more autonomy, they also need to govern those systems appropriately. Agentic AI systems are probabilistic, meaning their responses and actions are not always predictable, while many traditional cybersecurity controls are designed to operate deterministically, following predefined rules. That mismatch creates new risks for network operations teams.

Consider a firewall rule that appears unnecessary to an AI system. The rule allows a particular IP address to access a database process for one day each quarter. An AI system might identify the rule as unnecessary and remove it without knowing that the rule supports a legitimate business process. The rule looks like a mistake. It looks like something that should be eliminated immediately. But it may be essential for a quarterly financial process, a compliance workflow, or a partner integration. Removing it could cause a disruption that is difficult to diagnose because the action was taken automatically.

This example illustrates why network teams must determine where human review remains necessary as enterprises allow AI to take on more operational functions. AI governance requires human oversight for unusual exceptions, business-critical rules, and potentially harmful or inaccurate outputs. Controls should be tailored specifically to an organization's risk profile. A highly regulated enterprise may require stricter approval workflows, while a less regulated organization may allow more autonomy for low-risk actions. The common thread is that AI systems need boundaries, monitoring, and accountability.

Operational implications for network teams

For network operations teams, the increasing use of AI and automation requires integration of observability data and workflows, stronger tool governance, and controls over what AI systems can and cannot do. Teams should not assume that adding AI to a fragmented tool stack will resolve the underlying operational problems. Instead, they should focus on reducing unnecessary complexity, standardizing data, defining ownership, and creating clear escalation paths for AI-driven actions.

The survey findings also suggest that organizations should evaluate AI adoption in the context of their governance maturity. Teams with centralized ownership and adequate resources are better positioned to benefit from AI because they have the processes and data needed to support it. Teams without those foundations may find that AI adds noise rather than clarity. The most effective approach is likely to be incremental: integrate the most critical data sources first, establish governance for the tools and workflows that matter most, and expand AI use cases as data quality and operational confidence improve.

Network operations leaders should also recognize that tool sprawl is not only a technology problem. It is an organizational problem that affects skills, staffing, incident response, and cost. Consolidation may help in some cases, but integration and governance are often more important than simply reducing the number of tools. The goal is to create a coherent operating environment in which data flows across domains and engineers can investigate problems without unnecessary friction.

As AI systems become more capable, the stakes around governance will rise. Organizations will need to decide which actions can be automated, which require human approval, and how to detect and correct errors when they occur. They will need to monitor AI outputs for accuracy and security risks, and they will need to ensure that AI does not undermine the controls that protect critical infrastructure. The survey results point to a future in which observability is not just about collecting more data or deploying more tools. It is about creating a governed, integrated, and trustworthy operational foundation that allows both humans and AI to make better decisions. The next phase of observability is not simply about adding AI to a crowded tool stack. It is about building the governance, data quality, and operational discipline that allow AI to act safely and usefully across increasingly complex infrastructure.

Source:Network World News