India generates nearly 20% of global data, but hosts only about 3% of the global data centre capacity (approximately 1.2-1.7 GW out of over 100 GW globally). It is estimated that 90% to 95% of data generated by Indians is stored or processed overseas. The tax holiday announced in Budget 2026 for foreign cloud service providers using Indian data centres is a timely and strategic intervention aimed at narrowing India’s data infrastructure gap. It is also a strong signal that India wants to become a serious global data centre and AI computing hub.
Power, approvals and connectivity remain key bottlenecks
However, industry players feel that tax incentives alone will not close the gap between the data India generates and the infrastructure it hosts. “Tax incentives are a strong start, but they are only one part of the equation,” said CP Gurnani, co-founder and vice chairman, AIONOS. Data centres, particularly those supporting AI workloads, are power intensive and sensitive to latency and reliability. Scaling capacity will require assured access to clean and affordable energy, faster land and environmental approvals, high-quality fibre connectivity, and consistent state-level policies.
Equally important is regulatory clarity around data localisation, cross-border data flows and ESG-related infrastructure standards, said Atul Puri, managing partner and co-founder, SW India. If fiscal incentives are matched with execution certainty and infrastructure readiness, India has the opportunity to move from being a large data generator to a global hub for cloud and AI computing.
Long-term certainty could anchor global AI infrastructure
“Despite generating nearly 20% of global data, India hosts only a small share of global data centre capacity. Long-term tax certainty until 2047 directly addresses a key hurdle for hyperscalers making large, irreversible investments,” Puri emphasised. By requiring greenfield development, domestic ownership of physical infrastructure, and Indian reseller participation, the framework ensures that global capital leads to local asset creation, employment, and stronger domestic digital ecosystems, rather than remaining an offshore extension of global networks.
By allowing the tax exemption till 2047, the government is signalling long-term intent and inviting global capital to build capacity early and at scale, said Sunil Gupta, co-founder, CEO and MD, Yotta Data Services. It recognises that cloud and data centres are no longer peripheral IT assets, but foundational infrastructure for AI, digital public services, and next-generation enterprise growth.
By insisting that these are greenfield facilities, anchored by Indian partners who own the physical infrastructure and serve domestic clients through Indian resellers, the policy smartly aligns global capital with local capability and control. It creates a win-win, said Gurnani. Global players get scale and market access, while Indian infrastructure providers, telcos and service partners participate meaningfully in the value chain. Crucially, this also nudges higher data localisation and builds the compute backbone required for India’s AI ambitions.
As AI workloads move from experimentation to regulated and mission-critical use cases, global cloud providers will rely on Indian operating entities and local infrastructure partners for compliance, execution, and scale, said Gupta. This creates a powerful opportunity for India to anchor global AI platforms within its ecosystem, while strengthening sovereign, AI-ready infrastructure built and operated domestically. “If executed well, the combination of policy stability, local partnerships, and infrastructure depth can position India not just as a consumer of global cloud services, but as a hub for global AI computing,” he added.
The proposed tax holiday is a strong and timely signal of India’s intent to attract global cloud and AI infrastructure investment, said Sharad Sanghi, co-founder and CEO of Neysa. To convert this momentum into long-term leadership, the next phase must focus on how AI-ready infrastructure is accessed and operated at scale. AI workloads place very different demands on performance, latency, and availability, which makes platform readiness as important as physical capacity.
“At Neysa, we are already seeing growing enterprise demand for AI workloads to be hosted and managed within India, driven by data sovereignty, regulatory clarity, and the need for predictable performance. Realising this opportunity will depend on how effectively compute, storage, networking, and AI software are orchestrated into secure, enterprise-grade platforms that sit above the underlying infrastructure,” Sanghi said.
As capacity expands, policy must also enable these AI cloud and platform layers so enterprises can train and deploy models efficiently within the country. “If infrastructure expansion and platform innovation move together, India can shift from being a large generator of data to a meaningful global destination for AI compute,” he added.