Introduction

Artificial Intelligence (AI) is rapidly expanding across almost every sector of the Indian economy—healthcare, agriculture, governance, education, finance, and industry. It is increasingly seen as a key driver of productivity, efficiency, and innovation, and features prominently in India’s vision of becoming a digital and knowledge-driven economy. At the same time, AI is also being promoted as a tool to tackle climate change, improve energy efficiency, and strengthen environmental monitoring. However, there is a major blind spot in this optimism: the environmental footprint of AI itself. The energy, water, and material resources required to build and run modern AI systems remain largely absent from policy debates. If India is to pursue AI at scale, it must also confront the ecological costs that come with it.

The Environmental Footprint of AI

The environmental impact of AI is not confined to a single stage of its use. It spans the entire lifecycle of AI systems—from training and deployment to maintenance and eventual disposal of hardware.

Energy Consumption

• Modern AI models, especially Large Language Models (LLMs), require enormous computational power.

• Training and running these models depends on:

  • High-performance data centres

  • Energy-intensive processors and GPUs

• Even routine AI queries consume significantly more energy than conventional digital searches or basic computing tasks.

• As AI adoption spreads across sectors, the cumulative energy demand of digital infrastructure is set to rise sharply.

Carbon Emissions

• Training a single large AI model can emit hundreds of thousands of kilograms of CO?, depending on the energy source used.

• The Information and Communication Technology (ICT) sector already accounts for a significant share of global greenhouse gas emissions.

• With the rapid growth of AI, this share is expected to increase unless:

  • Energy efficiency improves drastically

  • The shift to renewable energy accelerates

Water and Resource Use

• Data centres require large quantities of water for cooling servers and maintaining optimal operating conditions.

• In water-stressed regions, this can:

  • Compete with local communities and agriculture

  • Add to existing ecological pressures

• AI also increases demand for:

  • Rare earths and critical minerals

  • Electronic components and hardware

• Mining, processing, and manufacturing of these materials have their own environmental and social costs.

Lifecycle Impact

• The environmental cost of AI is not a one-time event.

• It includes:

  • Energy and emissions during training

  • Continuous power and cooling during deployment

  • Periodic hardware upgrades

  • Disposal of electronic waste at the end of equipment life

• Without a lifecycle perspective, the true ecological footprint of AI remains underestimated.

Global Recognition of the Problem

• At the international level, there is growing acknowledgment that AI has environmental costs.

• UNESCO’s 2021 Recommendation on the Ethics of AI explicitly recognises the negative impacts of AI on the environment and calls for sustainability to be part of AI governance.

• Studies by the United Nations Environment Programme (UNEP) have highlighted:

  • The rising energy and water stress caused by expanding digital and AI infrastructure

• However, global policy frameworks are still at an early stage, and binding standards remain limited.

The Indian Scenario: A Policy Blind Spot

• In India, the dominant AI policy discourse focuses on:

  • Economic growth and productivity

  • Digital public infrastructure

  • “AI for good” and “AI for climate solutions”

• While these are important, there is very little attention to:

  • The carbon footprint of AI systems

  • The water and energy usage of data centres

  • The material and waste footprint of digital infrastructure

• India’s existing Environmental Impact Assessment (EIA) framework:

  • Does not explicitly cover large-scale AI systems or data-intensive digital infrastructure as a distinct category

• As a result, a massive expansion of energy- and resource-intensive AI capacity could take place:

  • With minimal environmental scrutiny

  • And without long-term sustainability planning

Challenges

• There are no standard, widely accepted metrics in India to measure:

  • AI-related energy consumption

  • Emissions

  • Water use

  • Land and material footprint

• Companies developing or deploying large AI models are:

  • Not required to disclose their environmental impact in a transparent and comparable manner

• Data on:

  • Energy use

  • Carbon emissions

  • Water consumption of data centres
    remains fragmented and often unavailable to the public.

• This creates a real risk that:

  • India will lock itself into an unchecked expansion of highly energy-intensive digital infrastructure

  • At a time when it is also struggling to meet climate and sustainability goals

Way Forward: Making AI Environmentally Responsible

1. Measurement and Standards

• India should develop national standards to:

  • Measure energy use, emissions, water consumption, and land impact of AI systems and data centres

• Without measurement, regulation and accountability are impossible.

2. Regulatory Integration

• The scope of the Environmental Impact Assessment (EIA) framework should be:

  • Expanded to explicitly include large data centres and large-scale AI deployments

• This would ensure:

  • Environmental scrutiny before major digital infrastructure projects are approved

3. Disclosure and Transparency

• Environmental impact reporting of AI systems should be:

  • Made mandatory under ESG (Environmental, Social, Governance) and sustainability frameworks

• Companies should disclose:

  • Energy sources

  • Carbon footprint

  • Water usage

  • Efficiency measures

4. Promoting Sustainable AI Practices

• Policy should encourage:

  • Energy-efficient algorithms and hardware

  • Use of renewable energy for data centres

  • Greater reliance on pre-trained and smaller models where feasible, instead of always building ever-larger models

• Efficiency and sufficiency should become as important as performance and scale.

5. Multi-Stakeholder Policy Making

• The governance of AI’s environmental impact should involve:

  • Government

  • Technology companies

  • Academic and policy think tanks

  • Civil society and environmental experts

• This is essential to balance innovation, economic growth, and ecological responsibility.

Conclusion

AI will undoubtedly play a central role in India’s development journey in the coming decades. But technological progress that ignores environmental limits is neither sustainable nor wise. The ecological footprint of AI—its energy use, carbon emissions, water consumption, and material demand—must become part of mainstream policy thinking, not an afterthought. India still has the opportunity to shape a path of sustainable AI, where digital transformation goes hand in hand with climate responsibility. The real question is not whether India should invest in AI, but whether it can do so in a way that respects planetary boundaries and secures a livable future.