A chess board overlaid on a map of India with semiconductor chips, rare earth minerals, and trade routes forming the pieces -- some controlled by hands outside the frame

By Ramachandran Rajeev Kumar — 2026-04-02

The Dependencies India Doesn't Talk About

India has played its hand with skill. But skill is not strategy -- and the dependencies forming beneath the surface will not wait for the next election cycle to resolve themselves.

By Ramachandran Rajeev Kumar

In March 2026, India inaugurated its first operational semiconductor assembly facility in Sanand, Gujarat. That same month, India's annual trade deficit with China crossed $102 billion for the fiscal year -- a record. The first headline made the front pages. The second appeared, briefly, in the business sections before being overtaken by cricket scores and election analysis.

These are not competing stories. They are the same story, told from opposite ends.

India runs the most ambitious digital governance infrastructure on the planet on imported silicon. Every Aadhaar authentication, every UPI transaction -- 21.7 billion in a single month in late 2025 -- every CoWIN verification runs on chips designed in California, fabricated in Taiwan, and packaged in Southeast Asia. India wrote the code. Someone else made the hardware it executes on.

In a world where semiconductors have become strategic weapons, that asymmetry is not a quirk. It is a vulnerability with a timer.

The dependencies India discusses

Some risks are acknowledged. India's crude oil import dependency -- approximately 85 per cent of consumption -- is a permanent fixture in economic planning. The defence import question, once a source of quiet embarrassment at nearly 70 per cent, has seen genuine progress: indigenous defence production rose from Rs 46,429 crore in FY2014-15 to Rs 1.54 lakh crore in FY2024-25, with approximately 65 per cent of military equipment now manufactured domestically.

The Atmanirbhar Bharat programme has delivered results in specific sectors. India became the leading smartphone exporter to the United States in 2025, a 240 per cent year-on-year surge driven primarily by Apple's manufacturing shift. Defence exports reached record levels. PLI scheme investments materialised.

These are real achievements. But they risk creating a narrative of acceleration that obscures the dependencies that actually threaten India's trajectory.

The semiconductor mirage

India's semiconductor ambitions are real. Their timeline is the problem.

The Kaynes Semicon assembly facility in Sanand is operational. The Tata-PSMC fabrication plant in Dholera promises its first 28nm chips by December 2026. Headlines treat these as proof that India has entered the semiconductor race.

The 28nm process node was commercialised by TSMC in 2011. Fifteen years ago. The chips powering the latest iPhones, AI inference systems, and military guidance platforms run at 3nm and below. India is entering the race at a distance the leaders covered a decade and a half ago.

Scale tells the same story. India's approved semiconductor investments total approximately $17.3 billion across ten facilities. For context, TSMC alone spent $36 billion on capital expenditure in a single year -- 2024. Intel spent $25 billion. Samsung spent $47 billion. India's entire national semiconductor programme would not cover a single year's budget of any one of its three competitors.

When TSMC was reportedly invited to build a fab in India, it declined without public explanation, choosing instead to expand in the United States, Japan, and Germany -- markets with mature semiconductor ecosystems including ultra-pure chemical suppliers, photolithography infrastructure, and decades of institutional knowledge.

A 28nm assembly line is a start. Confusing a start with arrival is the kind of miscalculation that compounds across decades.

The invisible chokepoint

India possesses the world's third-largest rare earth reserves -- an estimated 6.9 million tonnes. It imports virtually all of its lithium, cobalt, nickel, and processed rare earth products. The reserves are in the ground. The processing capacity belongs to someone else.

China controls more than 60 per cent of global rare earth production and a larger share of processing and refining. In April 2025, Beijing imposed export controls on seven heavy rare earth elements. By October, the controls expanded to holmium, erbium, thulium, europium, and ytterbium -- reaching into components and assemblies manufactured anywhere in the world using Chinese-sourced materials.

These are not obscure elements. Lithium and cobalt power every battery in every electric vehicle and grid storage system -- central to India's 500 GW non-fossil fuel capacity target for 2030. Gallium and germanium are essential for semiconductor fabrication. Rare earth elements are irreplaceable in missile guidance systems, radar, fighter aircraft, and unmanned combat vehicles.

India's solar panel manufacturing -- promoted as a self-reliance success story -- discovered its supply-chain fragility overnight when China tightened export controls in 2025. The Union Budget 2026-27 repositioned critical minerals as a strategic pillar. Policy intent and processing infrastructure are separated by years of investment that has not yet begun at scale.

The AI-nuclear nexus

The Pahalgam crisis of 2025 compressed political reaction times as AI-enabled decision-support systems, autonomous drones, and algorithmic cyber operations entered the India-Pakistan operational theatre. Research published in early 2026 demonstrated that AI models placed in simulated military scenarios consistently escalated toward nuclear conflict -- a finding that should concern any nation in a nuclear neighbourhood.

India has articulated a framework for responsible military AI, emphasising human control over nuclear decisions. The deeper question remains unanswered: is India prepared to treat artificial intelligence not as a technology to be adopted incrementally but as a foundational pillar of national security architecture requiring dedicated infrastructure, talent pipelines, and governance frameworks built across decades?

The infrastructure alone demands staggering investment. Analysis projects India needing to pair AI data centre expansion with nuclear power in a convergence worth an estimated $200 billion. The talent pipeline remains thin, with the best AI researchers gravitating toward private-sector opportunities abroad.

The global lesson India cannot afford to ignore

What makes India's position particularly precarious is the timing.

The global architecture of interstate dependencies is fracturing in real time. On April 1, 2026, the President of the United States called NATO a "paper tiger" and stated that withdrawal was "beyond reconsideration." Five days earlier, reports emerged that Article 5 collective defence protections were being reframed as conditional on spending targets. The United States halted military aid to Ukraine by 99 per cent. It withdrew from 66 international organisations. The EU, which outsourced its security for three generations, is now spending hundreds of billions of euros to build military capacity it assumed it would never need.

This is not a distant drama. It is a structural lesson with direct application to India.

India has navigated the US-Russia-China triangle with tactical brilliance -- maintaining defence procurement from Moscow while deepening strategic partnerships with Washington, securing energy from sanctioned producers while expanding Quad cooperation. This multi-alignment has been celebrated, rightly, as diplomatic skill.

But diplomatic skill operates within a structure. And the structure is shifting.

If the United States -- which wrote NATO's founding treaty -- can walk away from a 77-year alliance, then the assumptions underpinning every Indian strategic partnership deserve stress-testing. The Quad is not a treaty alliance. AUKUS excludes India. The India-US defence cooperation framework depends on political continuity in Washington that recent events demonstrate cannot be assumed.

India's multi-alignment works because the current power distribution gives India options. The question is whether those options survive a world in which dependencies are weaponised -- where rare earth exports become leverage, semiconductor access becomes conditional, and intelligence-sharing partnerships are recalibrated with every election cycle.

The answer is: not without planning that extends far beyond the next electoral mandate.

What telescopic planning demands

Strategic foresight is not prediction. It is the systematic mapping of multiple plausible futures and the identification of decisions that remain sound across scenarios.

India's tactical tradition favours crisis response over horizon scanning. The country has demonstrated formidable ability to manage emergencies -- the Ukraine balance, the China border standoff, energy security under sanctions. This agility is a genuine national strength.

Agility without foresight is survival. It is not strategy.

The dependencies India faces are not five-year problems:

At five years: Accelerate the semiconductor pipeline beyond 28nm. Secure long-term critical mineral supply agreements with Australia, Chile, and African nations possessing unprocessed reserves. Build domestic AI research capacity through a national laboratory network with compensation that stops the talent drain.

At ten years: Achieve processing self-sufficiency in at least three critical mineral categories. Field indigenous AI-enabled defence systems tested against adversarial AI scenarios. Establish India as a trusted node in global semiconductor supply chains -- not at the leading edge, but essential to the chain.

At twenty years: Develop institutional capacity for technology governance capable of regulating AI systems India did not design. Create economic structures resilient to trade weaponisation from any single partner. Complete the energy transition infrastructure that eliminates crude oil import dependency as a strategic vulnerability.

At fifty years: Position India as one of three or four civilisational poles in a multipolar order -- through technological sovereignty, demographic resilience, and the credibility that comes from a society that plans beyond election cycles.

The demographic clock

One variable makes all of this urgent in a way that cannot be postponed.

India's working-age population advantage -- the demographic dividend -- peaks around 2040 and then begins to narrow. The window in which India has a structural labour advantage over ageing competitors -- China, Japan, Europe, and eventually the United States -- is approximately fifteen years. Every year without a strategic foresight architecture is a year in which the dividend depreciates without being deployed.

A young population that is educated, skilled, and employed is a civilisational asset. A young population that is underskilled and underemployed is a source of instability. The difference between these outcomes is not talent -- India produces more engineers per year than most nations produce graduates. The difference is planning.

The cost of not talking about it

The nations that plan across these horizons -- China with its Five-Year Plans and 2049 centenary vision, the United States with its industrial policy revival, Singapore and the UAE with their decades-long transformation roadmaps -- are not planning because they are wealthy. They are wealthy, in part, because they plan.

India's rise is conditional. It is conditional on decisions made in the next five to ten years about investments that mature over twenty to fifty. Every year without a strategic foresight framework at the centre of national policy -- not at its margins, not in think-tank white papers, but at the centre -- is a year in which dependencies deepen, optionality narrows, and the gap between India's ambitions and its structural preparedness widens.

India does not lack capability. It does not lack talent. It does not lack ambition.

What it lacks is the willingness to talk publicly, honestly, and without political varnish about the dependencies that could turn every one of those advantages into liabilities.

Strategic foresight is not a luxury for nations that have already arrived. It is the architecture of arrival itself.

Ramachandran Rajeev Kumar is Chief Executive of Aarksee Group of Companies, a Saudi Arabia-based conglomerate operating across carbon markets, green sciences, technology, and media. He writes on geopolitics and strategic foresight for BarathVector.

Sources: