Indo-US Working Group on Technology Policy: Data Center Build-Out and Data Sovereignty

This session of the Indo-US Working Group on Technology Policy examined the rapid expansion of data center infrastructure in India and the United States, focusing on the energy, permitting, supply chain, connectivity, and data sovereignty challenges that could shape future investment.

A central theme was that the principal constraints on data center development increasingly extend beyond the facilities themselves. In the United States, participants highlighted the mismatch between the speed of data center construction and the longer timelines required to expand grid capacity and complete interconnections. Aging grid infrastructure, constrained transmission capacity, and uncertainty over how infrastructure upgrades will be financed are becoming important bottlenecks. Participants noted that data center operators are increasingly considering on-site generation and a range of energy sources, including natural gas, solar, wind, and batteries, although no single clean-energy technology currently provides a complete solution for the reliability and scale required by large AI workloads. Water use and cooling technologies were also discussed, with newer closed-loop and other approaches potentially reducing consumption.

Participants emphasized that local opposition can become a significant obstacle even where national and state governments support data center investment. In the United States, concerns about energy, water, and other local impacts have contributed to moratoriums and debates over state preemption and local authority. Participants noted that the association between data centers and AI can amplify public skepticism. Transparency around resource use, infrastructure requirements, and economic benefits was identified as one way to narrow this trust gap. Participants also stressed that workforce policy should encompass not only construction and data center employment but the broader ecosystem of jobs and economic activity enabled by the infrastructure.

Similar challenges are emerging in India, although the political environment remains broadly supportive of data center investment. Participants noted that India's data center capacity remains concentrated in a relatively small number of metropolitan areas, creating pressure on local power and water resources. Transmission infrastructure was identified as an especially immediate constraint: the pace of investment is outstripping the ability of utilities and substations to connect new facilities. Participants suggested that more advance information from investors and greater geographic distribution of facilities could help infrastructure providers plan for demand. While protests and community concerns have emerged around several projects, participants indicated that these concerns have not yet generally stalled construction.

The discussion also highlighted the importance of clearer definitions of data and technological sovereignty. Participants noted that sovereignty increasingly encompasses not only where data is stored, but also who can access it, which laws govern it, and who controls the underlying cloud, chips, compute, and AI models. Simply locating a data center within a country's borders therefore does not necessarily provide the degree of legal or operational control that policymakers may seek. Participants cautioned against pursuing complete domestic technological self-sufficiency, arguing instead for trusted technology relationships and technical controls that can provide meaningful operational control while preserving access to global technology ecosystems.

For India, participants discussed the potential tension between domestic sovereignty objectives and reliance on U.S. technology providers. India's emerging cloud sovereignty framework may include nationality requirements for providers seeking to host certain sensitive categories of data, while further discussions around model sovereignty are expected to address how foreign AI models are deployed in India. Participants suggested that greater clarity between physical data residency and jurisdictional control could help address these concerns. One potential avenue discussed was a bilateral data access arrangement, potentially including tiered approaches for especially sensitive data and the use of technologies such as encryption and hardware security modules to preserve customer control.

Participants identified several areas where U.S.-India cooperation could help support the build-out. These included streamlining customs and certification procedures for data center equipment, improving supply-chain coordination for GPUs, cooling systems, and server infrastructure, and exploring development-finance mechanisms to support grid and clean-energy investments. The broader supply chain also emerged as a concern, including growing dependence on a small number of global memory-chip suppliers and the potential implications of memory shortages for both data center construction and consumer devices through which many Indians will access AI.

Overall, the discussion suggested that the U.S.-India data center relationship is moving beyond a simple question of attracting investment. Both countries face the challenge of building infrastructure quickly while addressing resource constraints, community concerns, grid limitations, and questions about technological control. Participants saw opportunities for cooperation around trusted infrastructure, resilient supply chains, energy investment, and clearer rules for cross-border data access that could allow both countries to expand AI capacity without treating sovereignty as synonymous with technological isolation.

The Indo-US Working Group on Technology Policy is a Track II dialogue convened by The Dialogue and the Information Technology and Innovation Foundation (ITIF). Session summaries do not attribute comments to individual participants.