What is the environmental impact of AI?

AI systems, particularly those using large language models, consume significant amounts of electricity and water, contributing to environmental concerns.

Why is the electricity demand from AI data centres increasing?

The rapid adoption of generative AI has led to an expansion in hyperscale data centres, increasing electricity demand.

What other resources do AI data centres consume?

AI data centres also consume water, particularly for cooling systems, which can impact regions with water stress.

Why is consistent reporting on data centre resource use important?

Consistent reporting is crucial for governments and communities to assess the environmental impact of AI expansion.

ChatGPT's Daily Prompts Use Enough Electricity to Power a Small City

Q: How much electricity does a typical ChatGPT prompt consume?

A typical ChatGPT prompt consumes an estimated 0.34 watt-hours of electricity.

Every time someone asks ChatGPT to write an email, explain a concept or generate code, the exchange feels almost effortless. Yet behind those few seconds of text generation lies an expanding network of data centres consuming enormous amounts of electricity, water, and computing power.

The environmental impact of artificial intelligence has come under increasing scrutiny as AI adoption accelerates worldwide. OpenAI says ChatGPT now processes around 2.5 billion prompts every day, highlighting the extraordinary scale at which modern AI systems now operate. As governments and technology companies race to build more powerful models, experts warn that the infrastructure powering those services is becoming one of the fastest-growing sources of electricity demand.

Billions of Small Prompts Add Up Quickly

According to research published by EARTHDAY.ORG, a typical ChatGPT prompt consumes an estimated 0.34 watt-hours of electricity, roughly enough energy to power a standard LED light bulb for around two minutes.

While that figure appears negligible in isolation, multiplying it by 2.5 billion daily prompts results in an estimated 850 megawatt-hours of electricity every day. That is enough electricity to power a small city or tens of thousands of homes for 24 hours, depending on local consumption patterns.

The comparison illustrates how billions of seemingly insignificant digital interactions can collectively create a substantial energy footprint.

Bloomberg Technology explains why soaring AI demand is driving record investment in data centres, electricity infrastructure, and computing capacity.

AI's Electricity Demand Is Rising Faster Than Ever

The issue extends well beyond ChatGPT itself. The rapid adoption of generative AI has triggered an unprecedented expansion in hyperscale data centres, which house the specialised processors needed to train and run large language models. Companies including Microsoft, Google, Amazon, and Meta are investing billions of pounds in new facilities, advanced chips, and power infrastructure as AI usage continues to grow.

According to the International Energy Agency (IEA), global electricity demand from data centres grew by 17% in 2025, while electricity consumption from AI-focused data centres surged 50% in a single year. The agency expects overall data centre electricity demand to roughly double from 485 terawatt-hours in 2025 to around 950 terawatt-hours by 2030, accounting for approximately 3% of global electricity demand.

The IEA also noted that the AI industry is creating growing competition for electricity supplies, grid connections, advanced semiconductors, and investment capital as companies race to expand computing capacity.

The Hidden Cost Goes Beyond Electricity

Electricity is only part of AI's environmental footprint. Modern data centres also require sophisticated cooling systems to prevent thousands of high-performance processors from overheating. Many facilities use water-based cooling, making freshwater another increasingly valuable resource.

EARTHDAY.ORG estimates that US data centres consumed around 17 billion gallons of water during 2023. Researchers have also warned that many new AI data centres are being built in regions already facing long-term water stress, raising concerns about sustainability as demand continues to increase.

Experts have also highlighted the lack of consistent reporting around data centre energy and water use, making it difficult for governments and local communities to fully assess the environmental impact of future AI expansion.

Why It Matters

Artificial intelligence is becoming embedded in everyday life, from workplace productivity and education to software development and online search. But the AI race is increasingly becoming a competition over infrastructure rather than software alone.

As demand for AI services continues to rise, technology companies are investing heavily in data centres, renewable energy, semiconductor manufacturing, and electricity networks to secure the computing power needed for future growth.

For consumers, a single ChatGPT prompt may seem insignificant. At global scale, however, billions of daily interactions are reshaping electricity demand, infrastructure investment, and environmental policy. The future of AI will depend not only on building smarter models, but also on whether the systems powering them can expand without placing unsustainable pressure on power grids, water supplies, and climate targets.