The evolution of intelligent data infrastructure: Rethinking data control in the AI era

Enterprise data infrastructure has evolved well beyond its traditional role. At NetApp, we’ve seen this shift firsthand as customers increasingly expect their data platforms not only to store and protect data, but also to understand it, move it across environments, and make it usable across applications, analytics platforms, and AI systems. This shift toward intelligent data infrastructure is enabling organizations to operate data environments that span clouds, applications, and increasingly complex data pipelines.

But as organizations accelerate their adoption of AI, a new challenge is emerging. AI systems, automated workflows, and service identities now interact directly with enterprise datasets. Data is no longer accessed only by well-understood applications or known human users. It is increasingly consumed by agents, models, pipelines, and autonomous systems that retrieve and process information dynamically.

In many ways, this is an extension of a familiar challenge. For years, organizations have had to determine which users and applications should be allowed to access specific data. The difference today is that the set of actors interacting with enterprise data is expanding rapidly, and many of those interactions are happening automatically by AI agents at machine speed.

This shift raises a fundamental question:

Who—or what—should be allowed to interact with enterprise data, and how can that interaction be understood and controlled in real time?

For years we focused on where data is stored. The next decade will be defined by who can use that data and how its use is controlled.

Many modern security and governance tools help organizations answer an important question: Where is sensitive data located? They scan environments, classify information, and provide visibility into where potential risk might exist. This visibility has become an important part of how organizations manage their data environments.

But the challenge organizations increasingly face is not simply identifying where sensitive data resides. It is understanding how that data is being accessed and used as those interactions occur.

In modern data environments, access patterns are constantly evolving. Applications retrieve data dynamically. Automated processes interact with shared datasets. AI-driven systems query enterprise knowledge sources as part of their workflows. These interactions occur continuously across distributed environments, making them difficult to observe with traditional monitoring.

As a result, many organizations can see where sensitive data exists but have far less clarity into how that data is being used. In this environment, dashboards and alerts alone do not change the outcome.

Visibility may help organizations understand where risk exists. But visibility without the ability to intervene does not change the risk.

If understanding how data is being used is becoming critical, the next question is where that understanding - and control - can realistically occur.

From our vantage point, working with customers, the answer increasingly points to the data layer itself.

At NetApp, we sit at the center of these interactions through the data infrastructure our customers rely on every day. We see which identities access data, which applications and services interact with it, and how those interactions evolve over time. As automated systems, service identities, and AI-driven workflows begin interacting directly with enterprise datasets, this vantage point becomes uniquely important.

Because of this position in the architecture, we can observe the relationships surrounding data—the connections between datasets, users, applications, services, and the activity occurring between them.

This perspective makes it possible to move beyond simply identifying where sensitive data exists. It allows organizations to begin understanding how that data is being used, which interactions are expected, and where unexpected or risky patterns may emerge.

In other words, the place where data is managed is increasingly becoming the place where its use can be understood and ultimately controlled.

NetApp is the leader in intelligent data infrastructure, and our customers are telling us directly that they need help. Their most sensitive unstructured data already lives on NetApp, and they are asking us to help them discover and classify it, govern and enforce access controls, and keep AI systems in check—right where the data lives.

That is exactly what we are building. At RSA this year, we will share an early look at the work we’ve been doing with our customers and partners.

An AI assistant needs access to enterprise data—documents, knowledge bases, runbooks, internal communications—to be useful. But within those same repositories, sensitive information often exists—intentional or otherwise.

In this scenario, an internal AI assistant begins accessing data it shouldn’t. The platform evaluates that access in real time, understands the relationship between the data, the identity, and the behavior, and intervenes before any sensitive information is exposed.

This is one example of a broader shift. Any legitimate identity—human, application, or AI—can access data in ways that introduce risk. Controlling that risk requires both visibility and enforcement at the point of data access. Because NetApp sits directly in the data path, we can do exactly that, extending this approach across a wide range of data security and governance use cases.

Organizations already operate complex ecosystems of security, governance, and data management tools. The goal is not to add another standalone system into an already crowded environment. Instead, we believe the infrastructure layer can serve as a point of convergence - working alongside existing tools and partners while providing unique visibility into how data is being accessed and used.

This is why our approach is designed to integrate with the broader ecosystem. Partners across the data security and AI governance landscape—including Cyera and Enkrypt.AI—help customers understand data sensitivity, AI risk, and governance requirements. Combined with the vantage point of the data layer, these insights move organizations from isolated visibility toward coordinated understanding and control.

More importantly, we see this work as the beginning of a broader conversation with customers. As organizations rethink how data is accessed and used across increasingly complex environments, their feedback will help shape the evolution of these capabilities.

At NetApp, we have seen enterprise data infrastructure evolve in step with how organizations use their data. For years, the focus was on where data is stored—ensuring it could be protected, moved, and accessed reliably across increasingly complex environments. More recently, we have worked with customers to build intelligent data infrastructure that helps them understand their data and make it usable across applications, analytics platforms, and emerging AI systems.

The next step in that evolution may be to understand and influence how that data is used.

As automated systems, applications, and AI models increasingly interact directly with enterprise data environments, organizations will need new ways to observe those interactions, understand them, and respond when something unexpected occurs.

At NetApp, we believe the data layer has a unique role to play in helping customers navigate this shift. Our goal is to explore these capabilities alongside customers and partners as part of the continuing evolution of intelligent data infrastructure.

In an AI-driven world, the question is no longer just where data lives—it’s how it is used, and how that use is understood and controlled.

Speak to our product team and explore more about what we’re up to.