OPINION | Economic Survey highlights a new vision for tech sovereignty in India

India must still determine how best to regulate the technology.

The Economic Survey’s observations on compute and AI portend potentially significant developments in the upcoming Budget. First, it flags concerns raised by the Financial Times about off-balance-sheet leverage in global AI investments, alongside skepticism expressed by IBM’s CEO regarding the financial viability of large-scale data center expansion. Second, it highlights that the demands data centers place on water, energy, and finance may be difficult to reconcile with India’s economic and existential realities, where steady access to these basic amenities is still not universal. Taken together, these signals suggest that Budget 2026 is unlikely to offer subsidies or major outlays for data centers. This presents a unique opportunity for India to reimagine its conception of technological sovereignty.

Concerns Over AI Investments

In a digital economy typified by cross-border and multi-directional data flows, technological sovereignty need not hinge on physical localisation, but on predictability. Until now, the idea of technological sovereignty in India has largely meant building as much as possible domestically, and forcing localization where it was not. Ratcheting up the number of home-grown data centers fed into this vision.

The goal of localisation was understandable. They were originally driven by a quest for seamless access to data for law enforcement, which was often stymied by an antiquated and broken international data-sharing regime under Mutual Legal Assistance Treaties. They continue to be reinforced by anti-globalist sentiments, both at home and internationally.

However, these sentiments render the quest for self-reliance incompatible with the nature of digital flows today. Digital information flows cannot be reduced to a zero-sum game, where data moving into or out of a country is viewed as a net loss or gain.

India's Digital Ecosystem and Global Synergies

Research has shown, for instance, that large digital platforms create significant opportunities for Indian MSMEs, in terms of both revenue generation and customer acquisition, domestically as well as internationally. The network through which a domestic firm reaches a global customer may be foreign, but the service provided is wholly Indian. Similar mutualism exists across the AI stack. While infrastructure and models may originate with international or domestic firms, Indian companies are the ones building applications on top of them.

Here again, research shows that firms derive value from multiplicity in AI infrastructure and are careful to avoid lock-in to any single architecture. Data gathered by the Esya Centre from start-ups indicates that AI developers routinely multi-home across clouds and models—an effective way to limit capital intensity while still capturing the benefits of advanced infrastructure.

Lessons for Start-up Ecosystem

These synergies are often ignored when technological sovereignty is reduced to a static ledger of national credits and debits, where outward digital flows are automatically treated as a loss to the national interest rather than as an enabler of economic value and global integration.

A reimagined conception of technological sovereignty for India must begin with trust. Businesses value certainty, even more so in a global environment marked by policy volatility and erratic decision-making. Singapore offers a useful illustration. Despite its small size, it has emerged as the second-largest innovation hub in Asia and ninth globally, attracting start-ups from around the world, including India.

A 2022 report by the Asian Development Bank notes that this influx has been driven by startup- and investor-friendly policies, alongside an unwavering commitment to the rule of law. Tax certainty—including the absence of capital gains tax on investments—combined with a co-investment model in which the government invested alongside venture funds, helped crowd in private capital. Importantly, this approach also allowed the state to leverage the market expertise and risk-assessment capabilities of private investors, who are often better positioned than government agencies to evaluate fast-moving technology sectors.

Proximity to capital, coupled with clear and credible rules of the road, in turn drew start-ups to Singapore’s ecosystem. There is also evidence to suggest that investors themselves encouraged firms to relocate or incorporate there, so they could benefit from the legal and institutional assurances offered by a stable jurisdiction.

India could orient itself in a similar direction. Doing so, however, requires resisting the temptation to dilute or reverse institutional and legal commitments once made. Episodes such as the Tiger Global dispute are one example, but more recent decisions—such as accelerating the compliance timeline for certain entities under the Digital Personal Data Protection Act from 18 months to 12—also raise concerns. Data protection frameworks are complex to implement, and most jurisdictions globally have allowed at least two years for compliance with comparable regimes.

Navigating AI Regulation with Trust and Predictability

In the context of AI, the Economic Survey rightly notes that India must still determine how best to regulate the technology. Here too, trust should serve as the lodestar. This implies avoiding unnecessary compliance burdens or reactionary rule-making—such as the advisory issued in March 2024 that sought to effectively impose a licensing regime on all generative AI systems.

Trust enables digital flows through a country, and these flows enable the capture of value. This should be the modern conception of technological sovereignty for India—one that complements, rather than conflicts with, the country’s strengths as a hub of technical talent, ingenuity, and robust digital connectivity.

(Meghna Balis the Director of the Esya Centre, a tech policy think tank.)

Views are personal and do not represent the stand of this publication.