The thirsty cloud: The hidden environmental cost of Artificial Intelligence

By O.P. Srivastava, Maker of the documentary, The Dirty Sky

Artificial Intelligence is often imagined as clean, virtual, and immaterial — a revolution happening inside algorithms, clouds, and screens. We picture AI as weightless code floating in digital space. But the reality is far more physical, industrial, and ecological.

Behind every AI query, every chatbot conversation, every generative image, and every predictive model lies a rapidly expanding global infrastructure of data centres — massive industrial facilities consuming water, power, land, and natural resources at an unprecedented scale.

The cloud is not in the sky.

It is on the Earth.

And it is growing fast.

The Illusion of the Clean Cloud

The AI boom has triggered a global race to build hyperscale data centres. According to the International Energy Agency (IEA), global data centres already consume around 1–1.5% of the world’s electricity, and this demand is projected to double by 2030, driven largely by AI workloads, cloud computing, and digital services. Some projections suggest AI infrastructure alone could soon consume more power than many mid-sized countries.

Every AI system depends on heavy physical systems:

High-density server farms
Massive cooling infrastructure
Continuous power supply
Redundant energy backups
Water-intensive thermal regulation

This is not digital minimalism.

It is industrial-scale expansion.

AI is not replacing factories — it is becoming one.

The Machines That Drink Water
One of the least discussed consequences of AI infrastructure is water consumption.

Data centres use vast quantities of water for cooling. Hyperscale facilities can consume millions of litres per day, comparable to the daily water usage of small towns. Large AI clusters can draw hundreds of millions of litres annually, often from local groundwater systems.

In water-stressed regions, this creates silent competition between:

Communities
Agriculture
Urban water supply
Local ecosystems
Industrial AI infrastructure

Across parts of India, the United States, Latin America, and Africa, data centres are increasingly being located in regions already facing groundwater stress, simply because land availability and grid access make them commercially attractive.

Cooling systems for AI are becoming invisible extractive industries — operating quietly, continuously, and largely without public scrutiny.

The Power Hunger of Intelligence

AI is energy-intensive by design.

Research has shown that training large AI models can generate carbon emissions comparable to the lifetime emissions of multiple vehicles. At scale, AI data clusters consume hundreds of megawatts, equivalent to powering entire cities.

The challenge is not only volume — it is the source of energy:

Many data centres still rely on fossil-fuel-based grids

Diesel generators are widely used as backup systems

Renewable integration is often offset-based, not real-time

Grid stress increases peak-load instability

In many developing economies, AI infrastructure is expanding faster than clean energy transitions, creating a paradox:

Artificial Intelligence running on fossil intelligence.

Air, Heat, and Hidden Emissions

Beyond power and water, data centres generate local environmental impacts that are rarely discussed in mainstream technology narratives:

Heat islands caused by thermal discharge
Air pollution from diesel backup systems
Noise pollution
Micro-climate changes
Indirect emissions from global supply chains

These effects rarely appear in AI sustainability reports or corporate ESG narratives. They do not feature in innovation showcases or branding campaigns. But communities living near these facilities experience them daily.

The Human Cost of Invisible Infrastructure

Technology is never abstract to those who live beside its infrastructure.
A farmer watching groundwater levels fall.
A local doctor observing rising heat stress and respiratory issues.
A power engineer struggling with grid overload.
A young coder proud of AI innovation but uneasy about its footprint.
An environmental activist tracking invisible resource flows.
AI is reshaping human landscapes — not just digital ones.

The Moral Paradox of Progress
AI promises efficiency, intelligence, optimisation, and sustainability.

But its foundation is built on extraction:

Extracted water
Extracted energy
Extracted minerals
Extracted land
Extracted ecosystems

We are using intelligence to solve planetary problems by building systems that intensify ecological pressure.
This is the moral paradox of modern technological progress.

The Regulatory Blind Spot

Globally, AI governance frameworks focus on:
Ethics
Bias
Safety
Data privacy
Algorithmic accountability

But environmental accountability remains largely absent.

There are no universal standards for:

Water-use transparency
Cooling efficiency disclosure
Local ecological impact reporting
Grid load accountability
Carbon footprint per AI operation

AI is being regulated socially and ethically — but not ecologically.

Why This Conversation Matters Now
AI is no longer a future technology. It is present infrastructure.

Decisions made in the next decade will lock in:

Energy systems
Urban planning models
Water allocation patterns
Climate impact trajectories
Environmental risk curves

This is not just a technology debate. It is a planetary design question.

Conclusion

The future will not judge us by how intelligent our machines became. It will judge us by how responsibly we built them.

Because:

The sky is not the only frontier we are polluting.
Intelligence without ecological memory becomes extraction.
The future will not ask how fast we built machines.
It will ask what they cost the Earth.

By O.P. Srivastava, Maker of the documentary The Dirty Sky 📧 [email protected]

Author Note
This article is part of a larger research and documentary initiative titled The Dirty Sky, which explores the hidden environmental footprint of modern technological systems — from space debris to data centres, from digital clouds to planetary costs. The project examines the ecological consequences of invisible infrastructures that power our modern world.

Call for Collaboration
If you are working on AI sustainability, green computing, energy systems, climate modelling, water governance, digital infrastructure, or environmental policy — your voice matters in this conversation. Researchers, experts, institutions, and collaborators are invited to engage, respond, and contribute to this evolving story and the documentary film project.