Dennis Kozak is the Chief Executive Officer at Ivanti, and is responsible for the company’s overall strategic direction and growth.
getty
For much of the past decade, visibility has been one of enterprise IT’s most persistent challenges. Teams could not secure, manage or remediate what they could not see, so they invested heavily in monitoring, scanning and dashboards.
As a result, visibility has improved. Yet, the harder problem is acting on all of that information quickly enough to reduce risk. I often see this frustration among customers: The issue is not a lack of data, but an inability to use it effectively when decisions need to be made quickly.
As organizations pursue AI, that challenge becomes even more significant. AI doesn’t reduce the need for trusted data—it increases it.
Fragmentation Has Become Too Risky
Beyond data and visibility, the larger issue is structure.
Enterprise IT has evolved around specialized tools—one or more for endpoints, one or more for service management, many for security and maybe one for asset tracking (to name a few). Each may perform its function well, but none provides a shared, trusted system of record.
As a result, data becomes scattered, and scattered data is slow to act on. Questions that should take seconds—what is this device, should it be able to connect to our network, who owns it, is it compliant and what depends on it?— become a scavenger hunt across systems that describe the same machine in multiple ways. In one of my company’s surveys, we found that 55% of organizations described their IT and security data as siloed. At scale, that fragmentation is what turns a fast fix into a slow one.
For years, organizations were willing to tolerate this inefficiency because people could compensate for it. Teams manually reconciled data, connected processes and filled operational gaps. But AI changes that equation.
The more organizations automate decisions and actions, the less room there is for conflicting information.
AI Is Forcing A New Conversation
I see patching as the clearest example of this. The volume of vulnerabilities being disclosed each month has surpassed what manual triage can reasonably manage, and AI-assisted discovery is only increasing the pressure.
What was once routine maintenance now carries material business risk. Boards and audit committees are asking about patch compliance directly—a topic that would have remained buried within IT a few years ago.
This is the new reality of enterprise technology: Operational decisions are becoming business decisions. When the failure mode is a data breach, patching is no longer an IT chore; it is an enterprise risk issue.
AI cannot independently determine what assets exist, who owns them, what changed or whether an action complies with policy. Those answers come from a trusted system of record.
From Systems Of Work To Systems Of Record
The common response to fragmentation is to add another tool to connect the tools already in place. Too often, that creates another layer of complexity rather than removing one.
I believe a stronger approach is to make integration the starting point. That means consolidating around a shared system of record: one trusted layer that understands what exists across the environment, how it is connected and what state it is in. It may be less visible than the newest AI feature, but it is what makes those AI capabilities effective.
This represents a broader shift occurring across enterprise IT. Organizations are moving beyond systems of work that track tasks and workflows and toward systems of record that provide authoritative context for decisions.
A trusted system of record understands what exists across the environment, how assets relate to one another, who owns them, what changed and what state they are currently in.
That context is what modern AI systems depend on, and it is required for IT and security teams to be able to move toward an autonomous operation.
Data Comes Before Autonomy
Autonomy is only as reliable as the data behind it. Many AI initiatives do not fail because the model is weak; they fail because the model is working from incomplete, outdated or conflicting information.
AI can process information and make recommendations faster than any human team. What it cannot do on its own is know what the organization owns, understand the current state of the environment or answer for compliance. In other words, AI cannot replace data authority. Only a trusted system of record can provide that foundation.
Autonomy also requires governance. Any system capable of taking action needs clear guardrails: AI should handle the routine work it is best suited for, while humans retain oversight where judgment, accountability and risk decisions are required. Speed matters, but exposure is accelerating too — and speed without control only compounds risk.
The future is not autonomous systems operating without people. It is governed autonomy—where AI acts within defined policies, trusted data and human-defined boundaries.
Changing The Foundation And A Call To Action
As can be seen, three forces are converging: intensifying cyber threats, pressure to show measurable AI returns and board expectations to improve efficiency without adding complexity.
Therefore, leaders can no longer treat security, operations, data and AI as separate conversations. They are all connected, and the foundation that supports them must be connected as well.
While this does not mean ripping out what works or expecting transformation to happen overnight, the direction is clear, and the cost of maintaining fragmentation is becoming harder to justify.
Knowing this, leaders can start by setting the tone from the top: bring data together, allow systems to safely handle routine work and give teams the capacity to focus on the problems that require human judgment.
Forbes Business Council is the foremost growth and networking organization for business owners and leaders. Do I qualify?
