India’s technology sector entered a new phase during the week of September 6-10, moving past early AI experimentation toward building the infrastructure and governance frameworks required for large-scale deployment. The week’s developments highlighted investments in physical infrastructure, enterprise adoption challenges, and regulatory adaptations as the industry confronted the practical demands of AI integration.
Infrastructure Investments Scale Up
Tata Consultancy Services (TCS) announced plans for a ₹70,000-crore AI data-centre campus in Hyderabad, developed by its subsidiary HyperVault in partnership with other firms. The 1GW facility is designed to support high-density GPU workloads, reflecting a race to establish India’s physical infrastructure for frontier AI. TCS also launched an Industrial Autonomy & Engineering Lab in Pune that demonstrates AI-led manufacturing using robotics, digital twins and industrial AI.
The “lights-out” model is designed to show how factories can increasingly operate with minimal human intervention, while allowing manufacturers to test autonomous systems before deploying them on live production lines. Concurrently, HCLTech invested ₹185 crore to launch a 40,000 sq ft Advanced Semiconductor Lab in Bengaluru, featuring 25,000 sq ft of cleanroom space for post-silicon engineering, testing, and failure analysis. These projects signal a strategic focus on semiconductor capabilities and data-localization needs.
According to a Financial Express report cited in the source, India’s colocation data-centre capacity has grown fivefold in five years to approximately 2GW, with projections of reaching 10GW by 2031. AI workloads and data-localization requirements are driving the expansion, but power, cooling, water availability and the availability of suitable infrastructure may emerge as constraints as AI workloads intensify.
Enterprise AI Adoption Accelerates
Swiggy adopted Snowflake as a unified data platform to manage analytics across its food delivery, Instamart, and Dineout services, with Apache Iceberg providing a central analytical layer. The company reported that its slowest data workflows improved by 90-96%, with some queries dropping from two hours to 15 minutes. The initiative shows that enterprise AI success increasingly depends on resolving data-layer inefficiencies first.
Wipro, meanwhile, stated that AI initiatives have generated productivity equivalent to 20,000 employees, with redeployed capacity not directly replacing workers. Over 100,000 of its employees have undergone advanced AI training as the company transitions to a “human-AI operating model.” Despite these gains, questions persist about whether productivity improvements will translate into revenue and margin growth.
Regulatory and Governance Challenges Emerge
National Payments Corporation of India (NPCI) is developing protocols to authenticate AI agents operating on the Unified Payments Interface (UPI). This could eventually permit autonomous agents to execute routine transactions, though liability and control mechanisms remain unresolved. Separately, ServiceNow reported that Indian enterprises increased AI spending by 119% year-over-year, with AI now accounting for 16.6% of average IT budgets and projected to reach 21.3% by 2027. However, only 22% of firms have formal AI testing, auditing, or risk-assessment processes, revealing a gap between investment and governance.
The absence of structured oversight becomes more concerning as AI adoption accelerates. Mahindra Finance, for instance, scaled its partnership with Sarvam to deploy voice AI agents handling over one crore calls across sales, collections, and employee engagement in 12 Indian languages—a move notable for transitioning sovereign, multilingual AI from pilots to core workflows. This deployment represents one of the earliest instances of such AI being integrated into a large-scale financial-services workflow.
Market Reactions and Strategic Partnerships
OpenAI also expanded its work with Samsung Electronics on next-generation chips, emphasizing the infrastructure race extending beyond models to memory and specialized silicon. In a separate leadership change, Whatfix appointed co-founder Vara Kumar as CEO, effective immediately, following the death of former co-founder Khadim Batti on 1 September. Kumar, who co-founded Whatfix with Batti in 2014 and has led product and research, will now oversee the company’s next phase of growth.
