Amazon’s fresh pledge to invest $35 billion in India over five years, coming days after Microsoft’s $17.5 billion commitment to cloud and AI infrastructure, signals that India has become the primary global arena where US tech capital, digital regulation, and industrial strategy intersect. Together with Google’s $15 billion AI data‑centre hub, these moves amount to nearly $70 billion of new announced big‑tech capex in India’s digital stack over the next half‑decade. The scale, sectoral focus, and timing suggest that what is playing out is not just a growth bet, but a strategic alignment between India’s state‑led “digital public infrastructure” ambitions and Washington’s desire to anchor the next wave of AI and cloud capacity in a friendly, populous democracy rather than in China.
The new numbers: capital, timelines, sectors
Amazon has raised its India pledge to more than $35 billion over roughly five years, taking its total planned spending in the country to around $75 billion when past and previously announced commitments are included. The new money is supposed to deepen its retail and logistics network, expand AWS data‑centre regions, and inject AI capabilities into its marketplace, exports, and SME digitisation programmes. Amazon is explicitly tying this to national goals: it talks of enabling AI tools for about 15 million small businesses, growing cumulative exports supported via its platform to $80 billion, and helping create several million jobs (direct, indirect, and seasonal) by 2030.
Amazon announced plans to invest over $35 billion in India to expand operations by enhancing AI capabilities and increasing exports https://t.co/lvqhsHT2Hc pic.twitter.com/WWiLKAyxlk
— Reuters (@Reuters) December 10, 2025
Microsoft, for its part, has announced $17.5 billion over four years—its largest investment in Asia—to build out cloud and AI infrastructure, including expanding existing regions in Chennai, Hyderabad and Pune and launching a new “India South Central” hyperscale region expected around mid‑2026. The company is bundling infrastructure with human‑capital commitments, promising AI skills training for roughly 20 million Indians by 2030 and emphasising “sovereign capabilities” and population‑scale AI diffusion after CEO Satya Nadella’s meeting with the Indian Prime Minister Narendra Modi in New Delhi. In parallel, Google has said it will spend $15 billion over five years on an AI data‑centre hub in Visakhapatnam, pitched as its largest AI facility outside the US, with gigawatt‑scale power needs and an undersea cable gateway that effectively hard‑wires India into the company’s global compute and connectivity backbone.
Major US tech AI investments in India
Why India, why now? Demand, de‑risking, and DPI
Three forces explain the clustering of these announcements.
First, India has become one of the world’s largest long‑run demand pools for digital services, with over a billion mobile connections, rapidly growing data consumption, and an under‑penetrated enterprise cloud market. Hyperscalers see a chance to lock in market share before domestic competitors and Chinese cloud players can gain meaningful footholds, especially as India’s data‑localisation and sectoral compliance rules make local capacity a quasi‑regulatory necessity. Second, Western companies are actively re‑routing capital away from China and into “China‑plus‑one” locations; India offers both geopolitical alignment with the US and the promise of scale, making it a target not just for manufacturing but for data‑centre and AI compute clusters. Third, New Delhi’s success in building digital public infrastructure—Aadhaar, UPI, DigiLocker, ONDC—has created a unique interplay of state rails and private trains: big tech can ride on top of public rails but is increasingly nudged to host and process data locally, creating structural demand for hyperscale investment.
Thank you, PM @narendramodi ji, for an inspiring conversation on India’s AI opportunity. To support the country’s ambitions, Microsoft is committing US$17.5B—our largest investment ever in Asia—to help build the infrastructure, skills, and sovereign capabilities needed for… pic.twitter.com/NdFEpWzoyZ
— Satya Nadella (@satyanadella) December 9, 2025
The political theatre around these pledges underscores their strategic character. Microsoft’s announcement followed Nadella’s direct engagement with Modi, framing the money as support for India’s “AI‑driven future” and sovereign capabilities. Google’s Visakhapatnam hub was unveiled in tandem with messaging about a “digital backbone” interlinking Indian regions and linking them to a global AI network. Amazon’s figure was unveiled at its Smbhav summit alongside an economic impact study emphasising jobs, exports, and the company’s position as one of India’s largest foreign investors. These are not atomised corporate decisions but coordinated signalling between a government pursuing techno‑nationalist growth and companies seeking regulatory goodwill and long‑term market access.
Jobs, exports and productivity: the promise and the caveats
On paper, the macro signals are impressive. Amazon’s own numbers talk of helping generate millions of jobs and boosting exports to $80 billion by 2030 via its seller base, on top of the roughly $40 billion it has cumulatively invested since 2010. Microsoft and Google’s data‑centre clusters imply thousands of direct high‑skill roles and a larger layer of construction, maintenance, and ancillary service jobs, while skills programmes aimed at tens of millions of Indians could, if executed seriously, improve employability across IT services, startups, and traditional sectors adopting AI.
Yet the employment story is more complex than press releases suggest. Hyperscale data centres and cloud platforms are capital‑intensive and relatively light on long‑term direct jobs; the real gains come from productivity spillovers when SMEs, public agencies, and large Indian firms use cheaper compute and AI tools to transform their own processes. That diffusion is neither automatic nor evenly distributed: regions with better power, connectivity, urban infrastructure and human capital—Hyderabad, Bengaluru, Chennai, Pune, NCR—will capture a disproportionate share, potentially widening intra‑India digital and income divides unless state governments align their own policies and education systems with this wave of investment.
Exports are another double‑edged narrative. Amazon’s promise of $80 billion in cumulative e‑commerce exports by 2030 suggests India as a base for MSME‑led cross‑border trade, with global demand mediated by a US‑owned platform. That may help many small manufacturers and artisans, but it also locks them into proprietary logistics, search and advertising systems where pricing power and data asymmetries favour the platform over suppliers, reinforcing long‑running antitrust and competition concerns already raised against big tech around the world.
Sovereignty, regulation and the “gatekeeper” dilemma
One striking commonality across the Microsoft and Google announcements is the language of “sovereign capabilities” and nation‑scale AI. Data‑centre regions in Hyderabad, Chennai, Pune and Visakhapatnam, paired with undersea cable gateways and domestic power generation, are being marketed as assets that strengthen India’s digital sovereignty: data stays locally, latency falls, domestic firms and the state get access to world‑class compute. Yet ownership, control of core software stacks, and the ability to unilaterally change terms of service remain with foreign corporations answerable to regulators and courts outside India as well.
This is where India’s regulatory choices over the next few years will matter more than the headline billions. Unlike the EU’s Digital Markets Act, which explicitly targets “gatekeepers”, India has adopted a patchwork of sectoral rules (from payments to e‑commerce) and a data‑protection law that critics argue leaves significant discretion to the executive. As Amazon, Microsoft and Google entrench themselves as infrastructural players—owning not just consumer platforms but the cloud on which Indian startups, media, fintech, and even government services run—the state must grapple with a structural conflict: it wants these firms’ capital and technology, but it also wants to prevent them from becoming unaccountable utilities that can shape everything from retail competition to political content flows.
The companies are clearly aware of this tension. Microsoft’s emphasis on supporting “sovereign capabilities” and working with Indian partners, Google’s framing of its AI hub as part of a global network rather than simply an export of US services, and Amazon’s stress on helping small businesses and job creation are all designed to politically derisk their presence. But the fundamental asymmetry—the code, core models and platform rules are theirs; the data, users and political accountability are India’s—remains unresolved and will intensify as AI systems move deeper into governance and public service delivery.
Climate, power, and the invisible externalities
Less visible in launch speeches is the material footprint of this AI build‑out. Google’s Visakhapatnam hub alone is planned with an initial 1‑gigawatt capacity, with scope to expand to “multiple gigawatts”, a scale comparable to several large thermal power plants. Microsoft and Amazon’s expansion of hyperscale regions across multiple Indian cities, combined with Meta and others’ existing data‑centre plans, will collectively add staggering baseload power demand in a grid still heavily dependent on coal, even as India races to add renewables.
This raises uncomfortable questions. If AI data‑centres increasingly consume the cleanest available power—because that is where corporate net‑zero strategies push them—other sectors may find themselves more tightly bound to coal and gas, effectively “exporting” green electrons to big tech while domestic industry and households shoulder more of the emissions burden. Conversely, if policy pushes the hyperscalers to co‑invest in renewables, grid upgrades and storage in exchange for land and regulatory facilitation, India could leverage their capital to accelerate its broader energy transition instead of just subsidising corporate decarbonisation narratives. The current announcements hint at “extensive energy infrastructure” but remain light on binding targets, leaving a crucial part of the bargain opaque.
A new phase of the US–India tech compact
Seen together, these investments mark a new phase in the US–India relationship where digital and AI infrastructure become as central as defence deals and semiconductor fabs. For Washington, anchoring the next wave of AI and cloud capacity in India strengthens a friendly node in a contested technology order, building economic interdependence that makes any drift towards China or Russia costlier. For New Delhi, the money validates its narrative of India as a “Vishwaguru” of digital public infrastructure and a key theatre of the AI race, while giving it leverage to demand jobs, skilling, and localisation from companies that cannot afford to be shut out of such a large market.
However, if India treats Amazon’s $35 billion and Microsoft’s $17.5 billion as ends in themselves rather than as bargaining chips to secure technology transfer, domestic R&D, open standards, and space for Indian cloud and AI firms, it will have traded immediate growth optics for long‑term dependency. The real strategic test is whether, when this five‑year wave of spending is over, India has not just more data‑centres and exports booked through foreign platforms, but also its own competitive AI companies, robust regulatory institutions, and an energy system reshaped—rather than strained—by the demands of hyperscale computing.
