India’s Focus on AI and Its Environmental Impact
Kartavya Desk Staff
Source: TH
Subject: Environment
Context: India is set to host the AI Impact Summit 2026 in New Delhi, where it will champion the “Planet Sutra”—a global mandate to ensure AI development aligns with resource efficiency and climate resilience.
About India’s Focus on AI and Its Environmental Impact:
• India is currently at a technological crossroads. While the IndiaAI Mission (2024) and the development of homegrown models like BharatGen (launched June 2025) aim for digital sovereignty, the physical infrastructure—data centers—is straining the nation’s resources.
Key Data & Facts:
• Energy Demand: Global ICT is responsible for up to 3.9% of global greenhouse gas emissions. In India, data centre capacity is projected to reach 2,073 MW by 2027, an 85% increase from 2025 levels.
• Carbon Footprint: Training a single large AI model can emit over 626,000 pounds of CO2, equivalent to the lifetime emissions of five cars.
• Resource Intensity: A ChatGPT query consumes 10 times more electricity than a standard Google search.
• India’s Status: India has a 59% AI adoption rate, yet 50% of its data centers are located in extremely water-stressed regions like Bengaluru and Mumbai.
How AI Impacts the Environment?
• Massive Electricity Consumption: AI models require continuous high-density power for training and inference.
E.g. In Mumbai, the surge in AI-driven data centres has led to concerns over the city’s reliance on coal-based power to meet the 1,100+ MW load.
• Severe Water Scarcity: Cooling systems in data centres drink billions of liters of water to prevent hardware from melting.
E.g. In Bengaluru, data centres consume over 26 million liters of water annually, even as the city faced its worst water crisis in April 2024.
• Electronic Waste (E-waste) Explosion: The rapid obsolescence of AI-specific hardware (like GPUs) accelerates the toxic waste stream.
E.g. India generated 1.6 million metric tons of e-waste in 2024, with only a small fraction being formally recycled through pioneers like Attero.
• Carbon Emissions from Training: The computational brute force needed to train Large Language Models (LLMs) has a massive carbon price tag.
E.g. The development of sovereign LLMs in 2025 using thousands of GPUs has increased the Scope 2 emissions of Indian tech hubs.
• Natural Resource Depletion: Manufacturing AI chips requires rare earth minerals and ultrapure water.
E.g. India’s push for semiconductor fabrication (India Semiconductor Mission) in 2025 is raising concerns about groundwater depletion in manufacturing zones.
Challenges to Countering the Environmental Impact of AI:
• The Data-opaque problem: Because firms are not legally required to publish AI-model-wise energy and water use, sustainability reports hide the true ecological cost, preventing regulators and citizens from holding data centres accountable.
• The infrastructure–cooling paradox: In India’s hot climate, cooling high-performance GPUs consumes nearly as much power as computing itself, so expanding AI capacity actually multiplies electricity and water demand instead of just adding it.
• Fragmented regulatory frameworks: India’s EIA system is built for factories and mines, not for cloud-based AI firms, allowing massive GPU clusters to operate without environmental clearance despite their heavy digital carbon footprint.
• Hardware lifecycle & e-waste gap: AI chips become obsolete in 2–3 years, but India lacks advanced recycling plants to extract rare minerals, pushing toxic AI hardware into informal scrapyards that pollute soil and water.
• Energy-grid dependency: AI data centres need 24×7 stable power, but since India’s base-load electricity still comes mostly from coal and diesel backups, their green claims collapse whenever renewable supply fluctuates.
Solutions: The Way Forward
Global Context
• Legislative Action: The US AI Environmental Impacts Act of 2024 and the EU’s CSRD framework now mandate that tech giants disclose water and energy usage.
• UNESCO Recommendations: Over 190 countries have adopted non-binding ethics that emphasize AI’s “negative impacts on the environment.”
India’s Strategy
• Expanding EIA Scope: The Environmental Impact Assessment (EIA) Notification 2006 should be amended to mandate clearances for data centres exceeding 5 MW.
• ESG Disclosures: The Ministry of Corporate Affairs and SEBI can mandate Carbon Usage Effectiveness (CUE) reporting for AI companies.
• Adopting “Green AI”: Shifting from Red AI (resource-heavy) to Green AI, which prioritizes energy-efficient, pre-trained models.
• Renewable Integration: Incentivizing data centres to use 100% renewable energy, similar to the Haryana Water Resource Atlas (2025) approach for resource mapping.
• Standardized Metrics: Establishing national standards for Power Usage Effectiveness (PUE) to move toward water-positive data centres.
Conclusion:
India must move beyond viewing AI only as a tool for economic growth and recognize it as a resource-intensive industry that requires strict regulation. By integrating environmental audits into the IndiaAI Mission, the country can lead the global south in sustainable innovation. Ultimately, the goal is Green AI—where technological progress does not come at the cost of the planet’s vital resources.
Q. Artificial intelligence can deepen existing gender inequalities if social structures remain unchanged. Examine this statement in the context of unpaid care work. Also assess its social implications. (10 M)