NetApp CEO explains why data, not AI models, will decide the winners of the AI era

Ever since generative AI burst into the mainstream, the race has largely been framed around who can build the most powerful large language model or deploy the largest number of AI chips.

Technology companies have invested billions of dollars in developing frontier AI models, while governments are racing to build sovereign AI infrastructure.

But beneath the excitement lies a quieter shift that could ultimately determine which companies emerge as winners.

According to NetApp CEO George Kurian, artificial intelligence models will increasingly become accessible to everyone. What will remain difficult to replicate is an organisation's own data.

"We've long held the view that, in addition to talent and understanding your business or your clients deeply, data becomes the competitive moat of an industry," Kurian told CNBC-TV18 in an exclusive interview.

AI models are becoming commodities

Over the past two years, businesses have gained access to an expanding range of AI models from companies such as OpenAI, Google, Anthropic and Meta. Open-source alternatives have also narrowed the gap with proprietary systems.

That means companies increasingly have access to similar underlying AI capabilities.

"Everybody has access to the same chips, everybody has access to the same models, and the models themselves are largely trained on the same data," Kurian said.

If every enterprise can buy the same computing infrastructure and license comparable AI models, simply adopting AI will not create a lasting competitive advantage.

Instead, differentiation will come from how organisations combine those models with their own proprietary information.

Proprietary data is the real moat

Every organisation generates data that competitors cannot easily replicate.

Banks accumulate decades of transaction histories and customer behaviour.

Hospitals build vast repositories of diagnostic images and clinical records.

Manufacturers collect operational and production data from factories.

Retailers understand consumer purchasing patterns across millions of transactions.

Kurian believes these proprietary datasets represent the most valuable asset organisations possess in the AI era.

"Your ability, as an organisation, to deeply understand your business, your clients and your employees using your data becomes the competitive moat."

Rather than asking which AI model to deploy, business leaders increasingly need to ask whether their data is organised, secure and accessible enough to generate meaningful AI insights.

Why enterprise AI begins with data

Generative AI systems are only as effective as the information they can access.

While public AI models have broad general knowledge, they know little about an individual company's customers, products, operations or internal processes.

Enterprise AI therefore depends on connecting models with high-quality internal data.

That is where data management companies such as NetApp position themselves.

Rather than building foundation models themselves, they focus on helping organisations store, govern, protect and retrieve enterprise data so that AI systems can use it securely and efficiently.

For many businesses, preparing data for AI may prove a much larger undertaking than deploying the AI model itself.

Real-world AI depends on trusted data

The importance of enterprise data becomes even clearer in sectors where accuracy directly affects people's lives.

Healthcare is one such example.

NetApp has worked with organisations such as Sankara Nethralaya, one of India's leading eye care institutions, to help manage and harness large volumes of medical data. AI applications built on trusted clinical datasets have the potential to assist doctors in analysing medical images, identifying diseases earlier and improving patient outcomes.

The AI model itself may be available to many healthcare providers, but the quality, scale and governance of the underlying clinical data often determine how useful those systems become in practice.

The same principle applies across industries—from financial services and manufacturing to scientific research.

Kurian also highlighted NetApp's work with the European Space Agency, where large-scale data management supports efforts to map the universe. Such projects rely on collaboration between technology providers, research organisations and cloud partners to extract insights from enormous datasets.

In each case, the competitive advantage comes less from the AI model and more from the ability to manage and use unique data effectively.

Partnerships matter because no company can do it alone

As AI ecosystems become more complex, companies increasingly rely on partnerships rather than building every capability internally.

NetApp has expanded collaborations with companies including Google Cloud and Nutanix to help customers integrate AI into enterprise environments.

Kurian believes this collaborative approach reflects the realities of modern data centers, where organisations already operate multi-vendor technology environments.

"We've always believed it's better to harness the power of collaboration and community to build a better outcome for clients," he said.

For enterprises, successful AI deployments are likely to depend on combining infrastructure, cloud platforms, security, governance and domain expertise rather than relying on a single technology provider.

Data governance is becoming as important as AI itself

As organisations use more proprietary information to train and deploy AI systems, questions around data residency, sovereignty and regulatory compliance are becoming increasingly important.

Kurian argues that companies need to move beyond simply storing data within national borders.

True data sovereignty, he says, means organisations should be able to use globally available AI innovations while remaining compliant with local regulations and maintaining full control over sensitive information.

That capability is becoming particularly important for governments, banks, healthcare providers and other highly regulated industries.

Winning the AI race will depend on who uses data best

Much of today's AI conversation focuses on which company has built the largest model or secured the most advanced chips.

Kurian believes that perspective overlooks the factor most likely to determine long-term success.

As AI models become cheaper, faster and more widely available, they will become increasingly interchangeable.

What cannot be easily copied is decades of proprietary enterprise knowledge captured in structured and unstructured data.

The companies that invest in organising, protecting and making sense of that data will be better positioned to build differentiated AI applications, improve customer experiences and make better business decisions.

The AI era may be powered by models, but its biggest winners are likely to be determined by data.

For enterprises planning their AI strategy, that could be the most important lesson of all.