Should we be concerned about the environmental impact of AI in general?
Yes we should. In 2023 alone, big tech firms used enough water to fill 45 Olympic-sized swimming pools every single day.
Is There an Environmental Cost?
Earth Day 2025 brought a splash of cold water for the tech world — quite literally. A new wave of reports has put the environmental cost of artificial intelligence (AI) firmly in the spotlight, with one stat in particular making headlines: in 2023 alone, big tech firms used enough water to fill 45 Olympic-sized swimming pools every single day.
Behind the headlines lies a growing concern: as AI models get more powerful, they also get more demanding — not just in terms of energy, but in cooling systems that rely heavily on water.
Let’s take a closer look.
Water, Watts, and the AI Boom
From 2020 to 2023, companies like Google, Microsoft, Apple and Meta collectively ramped up their water usage by 60%. In total, they gulped down over 41 million cubic metres of water in 2023. To break that down:
- Google led the pack, using 59% of the total
- Microsoft followed with 19%
- Apple came in at 15%
- Meta rounded it out with 7%
Much of this is down to the cooling needs of AI hardware. Training and running large models takes a lot of computing muscle — and all that heat has to go somewhere. Many data centres rely on evaporative cooling systems, which use large volumes of water to keep temperatures down.
Is AI the Worst Offender in terms of Environmental Impact?
Not quite — but it’s fast climbing the ranks.
Let’s compare AI’s environmental footprint with some of its digital cousins:
- Data Centres in General
Already account for 4% of U.S. electricity use, rising to 6% by 2026. AI workloads add extra pressure, particularly on cooling systems. - Cloud Computing
Cloud services have big infrastructure needs, but AI’s use of high-performance GPUs makes it even more energy-hungry than typical cloud applications. - Consumer Tech
Think phones and laptops — these contribute heavily to e-waste, but AI’s main issue is less about landfill and more about power and water. - Cryptocurrency Mining
Still the heavyweight champ when it comes to electricity use — Bitcoin alone consumes more energy than some entire countries. AI hasn’t hit those heights (yet), but the trend line is rising.
What’s Being Done About the Environmental Impact of AI?
It’s not all doom and gloom. Although just now the environmental impact of AI appears to be pretty grim, the same tech industry that created the problem is now trying to fix it — or at least ease the load:
- Energy-efficient chips are in development
- Renewable-powered data centres are becoming more common
- AI is even being used to optimise energy use across other sectors, from logistics to agriculture
In other words, AI might be part of the problem — but it might also be part of the solution.
Why Environmental Impact Matters
This isn’t just about tech headlines. It’s about the real-world consequences of the digital tools we increasingly rely on — from AI chatbots to recommendation engines. As AI becomes more embedded in daily life, we need to ask tough questions about its sustainability.
Can innovation and conservation go hand in hand?
Let’s hope the answer is yes — for the sake of our planet as much as our progress.
Sources: