Whalesbook
The Infrastructure Surge: India's Data Center Expansion
India is experiencing an unprecedented build-out of digital infrastructure, positioning itself as a significant node in the global AI economy. Multi-billion-dollar investments are rapidly expanding hyperscale data centers, driven by policy incentives and a burgeoning digital market. By the end of 2025, India's total data center capacity reached approximately 1,700 MW, with projections indicating a 30% surge in 2026, adding an estimated 500 MW of new supply [3, 4]. Cumulative investment commitments in the sector have reached $126 billion by the close of 2025 and are forecast to exceed $180 billion in 2026, representing a 45% year-on-year increase [3, 4]. The market value of India's data center infrastructure is projected to reach $10.8 billion in 2026 and is expected to grow to $36.6 billion by 2035 [5]. This expansion is increasingly extending beyond major metropolitan hubs to tier-II cities, accommodating the demand for lower latency and data localization [3]. Global giants like Microsoft and Amazon have pledged over $50 billion for India's cloud and AI infrastructure, while Google is investing $15 billion in its largest data center hub outside the U.S. [38]. This infrastructure push has led to substantial investor interest, with data centers becoming a preferred asset class for regional investors anticipating significant price appreciation [3].
The Innovation Deficit: Lagging in Core AI Development
Despite the impressive growth in digital infrastructure, India's progress in domestic AI innovation remains notably constrained. While the country boasts a vast pool of technology professionals and leads globally in AI skill penetration and talent acquisition [2, 36], it lags significantly in creating foundational AI technologies and intellectual property. Between 2017 and 2024, India's share of global AI patent grants was negligible, hovering around 0.33% to 0.4%, a stark contrast to China's 64.7% and the US's 18.3% [6]. India's AI patent grant ratio stands at a mere 0.37%, considerably lower than its global patent application numbers which reached approximately 26,000 in 2024, placing it fourth globally in applications, though not grants [16, 21].
Research output also reflects this disparity. India ranks 14th globally in AI research paper contributions to top conferences, holding a 1.4% share between 2018-2023, far behind the US (30.4%) and China (22.8%) [29]. While India published over 17,000 AI papers in 2023 [17], its R&D expenditure as a percentage of GDP remains low at approximately 0.64%, significantly below innovation leaders like the US (3.47%) and China (2.41%) [10, 20]. The private sector's contribution to R&D in India is also notably low, accounting for roughly 36.4% of gross expenditure, compared to 75-77% in China and the US [10, 22]. Furthermore, India's compute capacity, measured in GPUs, trails far behind the US and China, with India planning to add 20,000 GPUs to its existing 38,000, while the US and China are expected to possess millions by late 2025 [32]. Private AI funding also highlights this gap, with India receiving $1.4 billion in 2023 compared to $67 billion in the US [32].
THE FORENSIC BEAR CASE: Value Capture and Structural Weaknesses
The core challenge for India lies in its ability to capture the full value generated by the AI revolution. The nation excels at providing the inputs—data, infrastructure, and engineering talent—but remains heavily reliant on external sources for high-value outputs like advanced models, platforms, and intellectual property. This pattern is mirrored within enterprises, where cloud adoption often translates to migrating legacy systems rather than fundamentally redesigning workflows for AI-driven automation. For example, real-time production dashboards might be implemented, but manual, slow approval processes persist, demonstrating that data flows instantly, but organizational evolution lags [cite: Rewritten News].
The talent paradox is another significant concern: India possesses a vast workforce with strong execution capabilities but has a limited base engaged in foundational AI research and development. The required deep research capabilities and advanced mathematics are less prevalent than implementation skills [cite: Rewritten News]. This leads to a situation where many AI startups act as intermediaries, building interfaces on global AI models, or engage in "AI-washing" rather than developing novel core technologies [cite: Rewritten News]. While innovative companies like Sarvam AI and Krutrim are developing indigenous language models [cite: Rewritten News], the broader ecosystem's focus remains on application rather than fundamental creation [cite: Rewritten News]. The AI patent grant ratio from academic institutions is particularly low at just 1% [16]. Moreover, the intermediate stage of R&D, from prototyping to market deployment (TRL 7-9), often represents a "valley of death" for Indian innovation, hindering the translation of research into scalable commercial products [20]. The country also faces challenges in retaining top AI talent and attracting foreign expertise, with a significant number of PhD holders pursuing opportunities abroad [14].
The Future Outlook
India's strategic focus on building robust digital infrastructure, including data centers and connectivity, provides a strong foundation for its AI ambitions. Policy initiatives like the IndiaAI Mission are actively promoting research and indigenous model development. Analysts are identifying companies poised to benefit from this infrastructure boom, including engineering, procurement, and construction firms like L&T, and power infrastructure providers such as Adani Green and NTPC, reflecting a market anticipation of continued development [24]. However, transitioning from merely hosting AI infrastructure to actively creating the intelligence that powers it will require a sustained focus on fostering deep-tech ventures and enterprise transformation. The real opportunity lies in leveraging India's scale, talent, and data to cultivate world-class AI innovation, moving beyond application and towards foundational creation to capture the full economic value of the AI age.