India Attracts $300 Billion in Data Centre Commitments, But Domestic Demand and Jobs Remain a Challenge: Report

India has attracted over $300 billion in data centre investment commitments within the past year, according to a recent report by US-based research and advisory firm Rhodium Group. Titled ‘Hyperscaler Heaven? A reality check on India’s AI ambition and geopolitical alignment’ and authored by Ankit Tiwari and Reva Goujon, the report assesses the country's infrastructure build-up, sovereign model strategy, and IT sector AI services.

While operational capacity is projected to triple by 2030, or potentially grow six-fold, the demand side of the ledger presents a different picture. The IndiaAI Mission’s subsidised graphics processing units (GPUs) are reportedly running at roughly 20 per cent utilisation, even though they are available below market rates. India’s largest GPU-as-a-service operator, Yotta, derives an estimated 80 per cent of its revenue from foreign clients and has agreed to distribute capacity through Nvidia’s global compute marketplace, DGX Cloud Lepton. Domestic enterprise AI workloads continue to route through global cloud providers rather than Indian servers, indicating that the newly built infrastructure is largely filled by overseas demand.

This reliance on external demand has implications for the political economy of the build-out. State governments initially justified incentive packages—such as land grants, power waivers, and capital subsidies—on the grounds of substantial employment generation. However, the Rhodium report documents a significant gap between these projections and actual outcomes. For instance, while Andhra Pradesh’s IT Minister cited potential for two lakh jobs from a single project, the state's own agency estimates only 200 direct operational positions. Construction employment peaks during the build phase and subsequently disperses. Consequently, states like Maharashtra, Tamil Nadu, and Karnataka have revised their green energy incentive frameworks.

On the geopolitical front, the report argues that India’s alignment has effectively been decided at the infrastructure level. Evidence includes India aligning its semiconductor export control list with US controls, adopting trusted-source requirements for telecom networks that exclude Chinese equipment, and taking positions in international spectrum forums that track the US-led coalition. In response, Beijing has reportedly recalled Chinese engineers from Foxconn plants in India, restricted tunnel boring machine exports, and selectively deployed rare earth controls.

In the realm of AI labs, initiatives backed by the IndiaAI Mission have focused on developing large language models, with Sarvam AI’s 105B parameter model emerging as a prominent result. While the Indic-language rationale is technically sound for processing queries in languages like Hindi, Tamil, or Bengali efficiently, the Rhodium report questions whether the domestic addressable market is large enough to fund frontier model development, noting that better-capitalised foreign labs are closing the performance gap.

Conversely, the report highlights the services layer as an area with substantial opportunity. TCS and Infosys reported annualised AI services revenue of 7.5 and 8 per cent of total revenue respectively, growing at double-digit rates amid a six-year low in overall IT revenue growth. Forward-deployed engineers—software professionals embedded within client organisations to customise AI deployments—are driving this growth. Infosys estimates the total addressable AI services market at $300 to $400 billion by 2030, leveraging decades of established enterprise relationships and deep context regarding legacy systems and workflows.

"The massive influx of capital into India's data centre infrastructure demonstrates strong global confidence in the region's digital foundation. However, as the Rhodium report points out, the real test for sustainable growth lies in bridging the gap between heavy infrastructure investments and domestic enterprise demand. For entrepreneurs and businesses operating in the AI and tech ecosystem, the immediate opportunity is not just in building massive hardware capacity, but in the services layer—specifically helping global and local enterprises integrate AI successfully into their existing workflows." — Dr. Shishir Gupta, Founder & CEO, StartupLanes