When 'AI Sovereignty' No Longer Equals 'Self-Sufficiency': India's Dilemma and the New Global Game

From "De-risking" to "Smart Interconnection": Redefining AI Sovereignty

In July 2026, Bloomberg launched a new monthly podcast series, *Emerging*. The debut episode focuses on India—the world's most populous country—as it seeks its own bearings in the shadow of the US-China AI duopoly. A conversation between host Menaka Doshi and Srikanth Velamakanni, co-founder of Fractal Analytics, reveals a concept that is being redefined: AI sovereignty.

For a long time, the term "sovereignty" in the technology sector was often equated with "self-sufficiency." Chips, operating systems, large models—nations rushed to build a complete closed loop from the bottom layer to applications. However, as technological complexity rises exponentially and the costs of global supply chain fragmentation grow increasingly high, this "full-stack" ambition is being corrected by reality. Velamakanni argues that AI sovereignty should not be an isolated infrastructure race, but an art of "smart interdependence."

India's "Middle Path" Dilemma

India boasts a vast digital population, a vibrant startup ecosystem, and global access advantages thanks to English. Yet, from infrastructure to cutting-edge algorithms, India has long been a follower in the AI arms race. The United States has giants like OpenAI, Google, and Meta; China has Baidu, ByteDance, and an ever-expanding computing network. India can neither replicate the capital and talent density of the US nor emulate China's government-led model.

Bloomberg's report accurately captures India's dilemma: on one hand, geopolitical pressures demand reduced reliance on certain countries; on the other hand, complete self-isolation will lead to technological gaps. India's solution may lie in building distinctive advantages in vertical domains—for example, using AI to optimize public services, agriculture, healthcare, and other essential local needs, while maintaining interoperability with international giants through open frameworks.

The Myth and Reality of Self-Sufficiency

Pursuing full self-sufficiency in AI is becoming increasingly impractical. Training a frontier large model requires billions of dollars in computing power, top-tier research talent, and a continuous flow of data. Even China remains constrained by export controls on high-end GPUs. Velamakanni's argument cuts to the core: no country can achieve independence across all layers of AI. The key is not to possess everything, but to control the most strategically critical nodes in the value chain.

This logic is reshaping the way national strategies are formulated. Japan, the European Union, and Southeast Asian countries are beginning to abandon the illusion of "omnipotence," shifting focus instead to data sovereignty, application-layer innovation, and the power to set standards. Sovereignty is increasingly reflected in the ability to manage data flows, the rule-making authority over algorithmic ethics, and guaranteed access to critical infrastructure—rather than physical "homegrown" production.

The Shape of a New Global AI OrderThe emergence of the "Emerging" podcast itself is a signal: emerging economies are redefining the future of globalization. In the past, economic sovereignty was often tied to resources and manufacturing capacity; now, data, algorithms, and computing power have become new strategic assets. The United States and China, as the first tier, are delineating spheres of influence through export controls, investment reviews, and talent competition. Other countries are forced to weigh between "choosing sides" and "multilateral maneuvering."

However, interdependence is not without risks. Over-embedding in another country's ecosystem can lead to technological dependence. Countries like India, Indonesia, and Brazil must draw a dynamic boundary between openness and security. Velamakanni's "smart interdependence" strategy suggests a middle path: maintain independent R&D in core sensitive areas, actively integrate into the global division of labor in non-core areas, and establish mutually recognized technical standards through multilateral agreements.

Long-term Trend: "Composite Sovereignty" in AI Governance

In the next decade, AI sovereignty will evolve into a composite concept. It includes not only computing autonomy at the hardware level but also algorithm sovereignty, data jurisdiction, and model safety norms at the software level. A country's competitiveness will no longer depend on the rate of technological localization, but on its position in the global AI network—whether it becomes a node or ends up as a terminal.

For an emerging giant like India, the real challenge may not be catching up with the US and China, but creating a new model of globalization: joining without dependence, cooperating while retaining the right to exit. While Bloomberg's report does not provide a definitive answer, it highlights the core issue of the era's transformation: when technological sovereignty is no longer a geographical concept, how should a country redefine its strategic domain?

This reflection applies not only to India but also to all countries seeking to safeguard their autonomy in the AI wave. After all, in the digital world, the strongest walls are often interconnected.