- AI is echoing the early fiber boom, with supply initially outstripping demand
- Simon Ninan argued enterprise AI adoption is being slowed by unclear ROI, data governance concerns and a lack of trust in AI outputs.
- The company sees opportunity in secure, on-prem and sovereign AI data foundations
NORMAN, OKLAHOMA – There’s an old saying that history doesn't repeat itself, but it often rhymes. That’s certainly the case if you look at today’s AI infrastructure boom and the fiber bonanza of the 1990s.
Simon Ninan, SVP of business strategy at Hitachi Vantara, compared the current AI capex cycle to the telecom industry’s heavy fiber investment decades ago. At the time, he said, operators laid vast amounts of fiber with a sort of ‘build it and they will come’ mentality. Years later, that capacity proved critical with the advent of the digital data era.
“That’s kind of the cycle that we’re in right now,” Ninan said. “We’re laying the fiber, we’re laying the pipes for AI. The demand needs to catch up.”
Put another way, Ninan believes the AI infrastructure boom of today isn’t necessarily a bubble story but more one of timing. The question is whether enterprises will move fast enough – and with enough production-scale use cases – to justify the spending already underway.
Ninan argued enterprise demand is still developing, in part because many companies have struggled to move AI pilots into production. He pointed to two barriers: unclear return on investment and a lack of trust in AI outputs.
The link between data and sovereignty
That trust gap is where Hitachi Vantara sees its opening. Ninan described the company’s AI role as providing “a trusted AI data foundation,” spanning how data is generated, protected, processed, transformed and ultimately used to produce business outcomes.
“Why does data governance matter? Because data governance is one of the top barriers right now to enterprise AI adoption,” he said. “When they don't trust that their systems and data can be protected from agents and from outside influences.”
The company’s strategy reflects an shift among enterprises toward more secure and controlled AI environments. Ninan said Hitachi Vantara expects these sovereign and on-premises data requirements to become increasingly important as companies assess the cybersecurity risks tied to AI models and sensitive enterprise data.
“While a lot of the initial use cases have been built out in hyperscale environments, we’re actually seeing a lot of enterprises place very high requirements on making data secure, bringing it on premises and putting boundaries around it,” Ninan said. “They may have done pilots with hyperscalers, but they will run their production in on premise and closed environments.”
Indeed, a recent study conducted by Omdia found 73% of enterprise respondents were actively piloting or deploying sovereign AI capabilities and 41% said they were allocating $1 million or more to sovereign AI over the next year. Sovereign AI was named as the highest strategic technology priority for 32% of respondents and it was in the top three for 81% of respondents.
“Regulatory pressure, geopolitical fragmentation and data localization requirements are forcing organizations to rethink how AI is built and deployed. At the same time, innovation can’t slow down,” Omdia Principal Analyst Mark Beccue said. "Sovereign AI is not a point solution. It's a system-level challenge."
If production workloads move closer to enterprise-controlled data, the opportunity will not just be in GPUs and cloud capacity, but in the underlying data pipelines, storage, security, reliability and networking needed to support AI at scale.
Scaling the vertical wall
Ninan connected Hitachi Vantara’s AI strategy to the broader Hitachi portfolio, which spans manufacturing, industrial systems, energy and transportation. As those sectors digitize and adopt AI, he said the opportunity is to connect business applications with trusted IT foundations in more vertically integrated ways.
“The future of IT is vertical,” Ninan said. “Not horizontal anymore, but vertical, and when we can create vertically integrated stacks that connect IT architectures to business outcomes, then we are delivering real ROI.”
Ninan added that it has increasingly noticed a fusion between business and IT stakeholders within customer organizations.
“The outcome that we have to generate for them is a business outcome enabled by IT,” he explained. “That's why this intersection, this vertical stack convergence, is such an important trend that I don't think people have fully appreciated and understood.”