Thirsty Giants: The AI Boom Drinking India's Water and Power

Question (10 marks, ~150 words):

"Any promise of jobs and prosperity from AI infrastructure must be measured against the land, power and water it consumes." Examine the opportunities and externalities of India's data-centre expansion, and suggest a regulatory framework that balances growth with sustainability and local interests.

Introduction:

• New York's first-in-the-US moratorium on large data centres (July 14) spotlights what India barely debates: its AI data-centre market is set to grow from ~$5.5 billion/1.6 GW to $90 billion/12-14 GW by 2035, in the most water-stressed of the top data-centre hosting nations.

Opportunities:

• Digital sovereignty: India hosts 20% of the world's data with ~3% of capacity; domestic compute underpins the AI mission, data-localisation goals and partnerships (Reliance-Meta, Adani-Google, Tata-OpenAI).
• Investment, ancillary construction activity and strategic depth in the AI value chain.

Externalities:

• Grid strain and groundwater depletion (a UP facility reportedly coincided with the water table falling from ~30 m to 180-250 m); few long-term jobs against large land, power and water footprints.
• Subsidy race (UP 2021 policy, TN, Telangana) socialises costs; debt-funded overbuild risks 'ghost infrastructure' if the AI cycle turns.

Regulatory framework:

• Tie incentives to enforceable water/energy norms and renewable power; mandatory resource-use disclosure; community-impact scrutiny before approval; green-quota models like Singapore's roadmap.

Way forward:

• Measure success by resource productivity and local benefit, not megawatts announced.

Conclusion:

• Regulating hyperscalers is not anti-AI — it is how India leads the Global South through the compute revolution sustainably.