Hidden Environmental Cost of India’s Digital and AI Boom: Data Centres, Water and Energy Security

Hidden Environmental Cost of India’s Digital and AI Boom: Data Centres, Water and Energy Security

(Based on article in Hindu Dated 9th August 2026)

Background and rapid expansion

• India’s digitalisation, cloud computing, 5G, e-commerce and especially Artificial Intelligence (AI) are producing unprecedented demand for computing infrastructure, making data centres a critical component of the country’s digital economy.
• Government data show that India’s data-centre capacity expanded four-fold—from about 375 MW in 2020 to over 1,500 MW by 2025. Mumbai/Navi Mumbai alone accounted for about 790 MW, followed by Chennai at 305 MW, Bengaluru at 182 MW and Hyderabad at 152 MW.
• The AI revolution will intensify this demand. Under the government's AI compute-capacity framework, about 38,231 GPUs through 14 empanelled service providers/data centres had been onboarded by March 2026.
• States are therefore competing for investment. Uttar Pradesh's earlier Data Centre Policy targeted 250 MW capacity and ₹20,000 crore investment, while the State has now notified a new Data Centre Policy, 2026.

Why data centres create an environmental challenge

• Huge electricity requirement: Servers operate continuously and AI workloads require enormous computing power; electricity is also required for cooling, networking and backup infrastructure.
• Water-intensive cooling: The uploaded article highlights an often overlooked “invisible water footprint” because conventional cooling systems can require large quantities of water to remove heat generated by servers.
• AI intensification: Generative-AI workloads involve high-density computing, meaning rapid expansion can increase both direct water consumption and indirect water use associated with electricity generation.
• Location problem: Data centres may be concentrated near major cities where land, electricity and connectivity are available, even though some such regions already experience serious water stress.
• Groundwater pressure: CGWB classifies areas with groundwater extraction exceeding annual extractable groundwater resources as over-exploited. Large new industrial consumers therefore require particularly careful scrutiny in water-stressed locations.
• Transparency deficit: The newspaper article raises concern that many data centres do not publicly disclose sufficiently granular information regarding their monthly water consumption, peak demand, water source and dependence on potable, groundwater or reclaimed water.

Solutions already emerging

• The Government says the industry is adopting direct-to-chip liquid cooling, adiabatic cooling, immersion cooling and high-density racks, which can improve water and energy efficiency for AI workloads.
• Groundwater extraction is regulated under Central Ground Water Authority guidelines. Certain large projects must submit hydrogeological impact assessments and groundwater-modelling studies before obtaining permission.
• Data centres should increasingly use treated wastewater/reclaimed water, closed-loop cooling and renewable electricity instead of competing with households for freshwater.
• Water-stressed regions could impose stronger environmental conditions, including mandatory Water Usage Effectiveness (WUE) and Power Usage Effectiveness (PUE) reporting.

Way Forward

• India needs a “Green Data Centre Framework” linking digital infrastructure policy with energy, groundwater and climate policy.
• Mandatory disclosure should cover annual and peak electricity consumption, freshwater withdrawal, source-wise water use, recycled-water percentage, carbon emissions and cooling efficiency.
• Environmental and water carrying capacity should become important criteria for selecting sites rather than allowing investment incentives alone to determine location.
• Renewable energy, battery storage, efficient processors, advanced cooling and wastewater recycling should progressively become standard features of hyperscale and AI data centres.
• Most importantly, India's ambition to become an AI and global data-centre hub must not create a conflict between digital growth and basic water security.

Conclusion: Data centres are indispensable infrastructure for Digital India and the AI economy, but their environmental footprint cannot remain invisible. The policy objective should be “more computing per unit of electricity and water”, enabling India to combine technological leadership with sustainable resource management.

Government references: MeitY/PIB — Data-centre capacity and cooling technologies · Ministry of Jal Shakti — Water consumption by data centres · CGWB — Groundwater regulation · UP Data Centre Policy 2026