Beyond Big Numbers: Making AI Investments Work for India’s Sovereignty

Major investment announcements by globaltechnology giants in India have understandably generated excitement across government, media, and the public. They are widely seen as a coming-of-age moment for India’s AI ecosystem—one that promises scale, capital, and global relevance. There is substance to this optimism. Yet, alongside celebration, there is a need for strategic clarity: what exactly does India gain, and what must it consciously retain—especially in terms of sovereignty, data control, and intellectual property?

It is worth recalling a moment from 2023, when Sam Altman, founder of OpenAI, remarked during his India visit that Indian startups attempting to build foundational AI models from scratch would find it “totally hopeless” to compete with OpenAI due to the scale of capital and compute required. The comment, perhaps arrogantly and clumsily phrased, nonetheless reflected a structural reality of AI economics. To his credit, Altman later acknowledged India’s strengths—its vast developer base, lower costs, policy support, and innovative potential—positioning India not as a passive consumer but as a serious AI stakeholder. 

Since then, investment commitments have followed at scale:

• Amazon: ~$35 billion by 2030
• Microsoft: ~$17.5 billion by 2029
• Google (Alphabet): ~$15 billion over the next few years

These are not trivial sums. They largely target AI, cloud infrastructure, and global capacity expansion—most visibly through hyperscale data centres.

Infrastructure Trade-offs

Data centres—often loosely referred to as “the cloud”—are the physical backbone of AI and other large data-oriented applications. Their migration to India is part of a global trend. In the US and Europe, environmental concerns around power consumption, water use, and land footprint have triggered growing resistance. Ireland, for instance, now sees data centres consuming close to a third of its electricity. Similar pressures exist elsewhere.

India, with its investment-friendly posture, growing digital demand, and improving infrastructure, is a natural destination. This presents opportunity—but also responsibility. Large AI data centres are energy-intensive, water-dependent, and require uninterrupted power. These are finite national resources, especially in a country where water stress and power deficits remain real constraints.

The question, therefore, is not whether such investments should come—but on what terms.

Value Shift

India’s IT story over the last three decades has been impressive, but also incomplete. Wage arbitrage and talent availability made India the world’s back office for software services. What we did not build, at scale, were globally dominant digital products—operating systems, productivity suites, platforms, or foundational AI models.

There are notable exceptions. Aadhaar, UPI, and India’s digital banking infrastructure stand out precisely because they combined government vision, policy backing, and sovereign ownership. These platforms are not merely successful—they are transformative.

By contrast, much of the work done in global tech firms’ India development centres involves writing code that ultimately strengthens foreign-owned products and IP. Indian engineers contribute enormously to global innovation, but ownership—and therefore long-term value—often resides elsewhere.

This is the pivotal point. If AI investments replicate the old services model, India risks remaining a high-end back office—this time for AI—without owning the engines, models, or platforms of the future.

Investment vs Capability

It is also important to be precise about economic returns. Data centres are capital-intensive but not employment-intensive. Much of the hardware—servers, GPUs, storage systems, networking equipment, and power conditioning units, probably even cooling fans—will be imported. Local jobs will largely involve maintenance and operations.

Crucially, hosting data centres in India does not automatically confer data ownership or control. Cloud architectures distribute data across geographies; physical location alone does not ensure sovereignty if the control rests elsewhere. Without clear policy frameworks, localisation of infrastructure may not translate into localisation of value or protection of Indian citizen’s data.

Strategic Imperative

None of this argues against welcoming global AI investment. On the contrary, India should actively attract it. But attraction must be paired with intentional design.

As India commits its power, water, land, and regulatory goodwill, it is reasonable—and necessary—to expect reciprocity in the form of:

• Co-ownership or clear rights over IPR generated by Indian teams working in India
• Data governance frameworks that ensure meaningful sovereign control
• Mandatory pathways for indigenous capability creation, not just service delivery
• Linkages with Indian industry, academia, and startups to build domestic AI stacks
• Long-term ecosystem commitments, not just infrastructure deployment

This is not protectionism. It is strategic nation-building.

Looking Ahead
AI will shape economic power, productivity, and national competitiveness for decades. India’s talent and market size ensure it will be a central arena in this transformation. The choice before policymakers is not between welcoming investment and asserting sovereignty—but between passive participation and purposeful leadership.

Big numbers make headlines. Ownership of ideas, models, and platforms will build powerful nations in coming years. India must ensure that the AI future being built on its soil also carries its imprint, its interests, and its intellectual signature.