Sovereign by design: Building the foundation for India’s AI future

Every major technology wave has arrived in India with the same promise: adopt it now, and the future is yours. In earlier waves, the country embraced transformative tools at remarkable speed but often built on foundations owned and controlled elsewhere. The convenience was real. So, over time, were the constraints. When the ground beneath a technology belongs to someone else, choices tend to narrow in ways that are difficult to see until they matter most. Artificial intelligence now offers India the chance to write a different ending, and that difference will be decided at the level of infrastructure.

The national ambition is unmistakable. National summits, policy momentum, and a growing consensus have positioned sovereign AI as the path from being a consumer of intelligence to becoming a global creator of it. But ambition has never been scaled on its own. What matters now is the foundation, and foundations cannot be retrofitted. Those who build sovereignty into the architecture from the start will lead; those who add it later will spend years playing catch-up, having lost the pace, resilience and strategic advantage that come from getting the foundation right.

What sovereignty really means

Sovereign AI is not isolation, and it is not a wall around the country’s data. It is control over how AI is designed, deployed, and managed, so regulatory compliance, and jurisdictional alignment sit alongside convenience and scale rather than being traded away for them. That control carries weight well beyond the technical. It shapes strategic autonomy, economic resilience, and dictates whether the benefits of AI reach citizens across the country as readily as they reach large enterprises in the metros.

The three pillars of India’s AI ambition

India’s sovereign AI ambitions rest on three pillars. The first is indigenous models, built for Indian languages, contexts, and the plurality of how the country actually communicates. The second is resilient domestic infrastructure, the compute, network, and storage backbone on which everything else depends, and the build-out is already accelerating. According to Gartner, India’s data centre systems segment is projected to grow 20.5% in 2026, the fastest-growing part of a national IT market expected to reach $176.3 billion, a clear signal that the physical foundation for sovereign AI is being laid in earnest. The third is stronger foundational research, the deep capability that keeps a nation building rather

than borrowing. Together, these pillars translate ambition into measurable impact across healthcare, education, agriculture, and financial inclusion.

Where national and enterprise needs meet

As India moves from vision to execution, two priorities come into focus, and the right foundation can serve both at once. A government wants AI systems it can trust—safe, reliable, and running on in-country infrastructure according to local regulations, so it can be confident that a model deployed in a hospital or a bank behaves safely and lawfully. An enterprise wants speed and flexibility to build and innovate without friction. These needs are often assumed to pull in opposite directions, but they do not have to. Both point to the same answer: trusted infrastructure deployed on-premises that delivers safety and speed together. Get the foundation right, and neither must be traded for the other.

Sovereignty begins at the foundation

This is why architecture matters more than rhetoric. When sovereignty is designed in at the network, compute, and storage layers, controlling how data moves, ensuring processing happens in trusted environments, and guaranteeing data residency at rest, alignment stops being a matter of negotiation and becomes a property of the system itself. Structural, not procedural. Verifiable, not assumed. Building it in from the start turns compliance from an ongoing burden into a built-in outcome, giving both policymakers and enterprises the assurance they need from the same design.

What the way forward looks like

That direction is already within reach. Hybrid cloud keeps sensitive and regulated data resident locally while still delivering the scale and elasticity AI demands, allowing organisations to test and innovate in-country without surrendering control of what matters most. Even as end-user public cloud spending in India is forecast to grow 28.1% to $17.5 billion in 2026 according to Gartner, enterprises are not abandoning local control but combining both, which is precisely the balance that trusted local infrastructure is built to serve. Systems that run securely and entirely in-country go further still, allowing AI to operate without depending on constant public cloud connectivity. Their quiet strength lies in the assurance they offer everyone: AI can be verified as safe, reliable, and compliant while everything stays within trusted local environments, so organisations can build with confidence rather than hesitation.

None of this is theoretical any longer. The capability exists today, and the decisions that will shape the next decade are being made now, in ministries, boardrooms, and data centres across the country. India already has the capital, the talent, and the ambition to become a global creator of AI rather than a perpetual consumer of it. What remains is to insist that the foundation be right, so that capability, control, and choice stay firmly in Indian hands. Sovereignty built in from the start is what will decide who leads the AI decade.

(The author is SVP and managing director, Hewlett Packard Enterprise, India. Views are personal.)