AI Infrastructure Boom Offers India a Growth Playbook as US Investment Reshapes Global Capital Flows

• AI infrastructure has become a major driver of US growth and dollar strength.

• US AI and data-centre spending may approach 2% of GDP in 2026.

• Big Tech capital expenditure is expected to exceed $700 billion this year.

• India is attracting global data-centre investment but needs faster execution.

• Power, land, connectivity, skills, and cybersecurity will be critical.

The United States has absorbed the heavy fiscal cost of the ongoing war in West Asia, estimated at about $40 billion, yet its economy has remained notably resilient. For many economists, the answer lies less in conventional explanations such as monetary policy or household consumption and more in the extraordinary investment surge around artificial intelligence.

AI has quickly become one of the most important capital spending stories in recent economic history. It is influencing growth, investment flows and even currency movements. For India, which cannot immediately match the American scale of spending, the US experience offers a strategy worth studying.

One familiar lesson in macroeconomics is the inverse relationship between the US dollar and precious metal prices. This matters for India because the country imports more than 90% of its gold, silver and copper. When commodity prices soften, they can partly cushion the impact of a stronger dollar on India’s current account deficit.

This raises a larger question: why has the dollar continued to strengthen despite wars, trade tensions and political uncertainty?

A major reason is investment. Over the past three years, the US has entered one of its biggest infrastructure spending cycles. But this time, the focus is not roads, bridges or railways. It is AI infrastructure.

What began in 2023 as excitement around consumer-facing AI tools has evolved into one of the largest concentrated capital expenditure booms in American history. The scale is being compared with the railroad expansion of the Gilded Age and the telecom buildout of the late 1990s. The AI data-centre boom is supporting US growth, strengthening the dollar and showing how digital infrastructure can become a macroeconomic force.

Until recently, investment in computing infrastructure attracted relatively little attention. Between 2015 and 2022, US spending on data centres, servers and networking equipment was barely 0.5% of GDP. Generative AI changed that dramatically.

By 2026, computing infrastructure’s share of nominal GDP had nearly tripled to around 1.6%, with most of the increase coming from AI-related hardware, data-centre construction and networking. Investment in AI data centres, hardware and networking reached about 1.4% of US GDP in the first quarter of 2026, up from roughly 0.7% in a short period. Including broader computing infrastructure, the figure rises to around 1.5% of GDP, more than double the average between 2015 and 2022.

The US is expected to spend close to 2% of GDP on AI and data-centre infrastructure this year, a level comparable to major public spending categories such as defence or education. It is also expected to account for more than 80% of the world’s estimated $800 billion AI infrastructure spending in 2026.

The impact is already visible in US growth data. Figures from the US Bureau of Economic

Analysis show that investment in information-processing equipment has become one of the strongest contributors to economic expansion. The Federal Reserve Bank of St Louis estimated that this category alone added 0.9 percentage points to real GDP growth in the first quarter of 2025.

By the first half of 2025, AI-linked investment had become the largest engine of US growth. Capital expenditure tied to AI contributed around 1.1 percentage points to GDP growth, overtaking consumer spending — a striking change in an economy usually driven by household consumption. The wider technology ecosystem added 2.28 percentage points to growth, around four times its contribution a year earlier.

Data-centre investment was especially significant, accounting for nearly 80% of the increase in final private domestic demand during the first half of 2025. Without this investment wave, US GDP growth would likely have been closer to 1% rather than 2%.

AI infrastructure is expensive because it requires massive upfront spending. Hyperscale technology companies such as Amazon, Microsoft, Alphabet, Meta and Oracle are expected to spend more than $700 billion in capital expenditure in 2026. Much of this will go toward graphics processors, servers, specialised chips, data centres and power infrastructure.

Such spending adds to GDP immediately through construction, equipment purchases and engineering work. The productivity gains from AI adoption may take longer to appear. The pattern resembles earlier technological revolutions, where infrastructure was built first and wider economic transformation followed gradually.

The Industrial Revolution offers a useful comparison. Thomas Edison’s work on electric light and power took place in the late 19th century, but many electrical appliances became common in American homes only decades later. AI may follow a similar path: the investment boom comes first, while the broader productivity payoff unfolds over time.

Critics argue that supportive monetary conditions may allow weaker companies to survive longer than they should. But over time, the process of creative destruction is likely to reshape industries as AI adoption spreads.

India cannot replicate the American investment surge overnight. With an economy of about $4 trillion, India is roughly one-eighth the size of the US economy. However, India’s digital infrastructure potential is already drawing international capital.

According to UNCTAD, data-centre projects helped lift India’s foreign direct investment by 44%, taking annual inflows to $39 billion. Alphabet’s proposed $15 billion AI hub and one-gigawatt data centre in Visakhapatnam shows the scale of investor interest. The project is equivalent to nearly 6% of Andhra Pradesh’s GDP. Meta is also building a major facility near Google’s proposed sites, while Amazon and Microsoft are investing in India’s digital infrastructure as well.

For India, the lesson is clear: it cannot afford to remain on the sidelines of the AI infrastructure race. But to benefit from this wave, the country will need more than investment announcements.

The first requirement is power. AI data centres consume enormous amounts of electricity, making reliable and affordable energy essential. Without sufficient power availability, large-scale AI infrastructure cannot function competitively.

The second requirement is execution speed. Land acquisition should not drag on for years, and approvals must move through predictable single-window systems. Investors need certainty, not prolonged procedural delays.

Connectivity is another priority. Stronger fibre networks and submarine cable landing stations will be needed to support the movement of massive volumes of data. Without deep digital connectivity, India’s data-centre ambitions will remain constrained.

Human capital will also matter. Universities must strengthen science, technology, engineering and mathematics education, while building curricula that support data science, AI engineering and related fields. A skilled workforce will be essential for operating and expanding this ecosystem.

Cybersecurity is equally important. Global AI investors will expect robust digital safeguards, resilient infrastructure and clear policy frameworks before committing long-term capital.

Countries that combine capital, infrastructure and regulatory certainty will capture the next wave of AI-linked investment. Those that fail to do so may end up as consumers of AI technologies rather than beneficiaries of the economic value they create.

India has a timely opportunity to position AI infrastructure as the base of its next investment cycle. To do that, it must treat AI not merely as another technology sector, but as a strategic economic platform capable of shaping growth, jobs, capital flows and competitiveness for years to come.