OPINION
Written by: Toral Veerasingham
Photo by: Abeer Arshad for The Fraser Post
Edited by: Katelyn Tsang
Designed and Formatted by : Mustafa S
AI is killing us, one data centre at a time

You wake up Monday morning at 8:05 and remember that you have an English paper due right when you get to school at 8:20. If this happened to you maybe 20-30 years ago your first instinct would be to jump out of bed and start madly scribbling on a piece of paper, hoping to get something down before you leave for school. But now, in the year 2025, all it takes is opening an app and writing a prompt, “Hey chat, write me a paragraph about the impacts of AI”, to get that assignment done. You got the assignment in on time, you didn’t even spend a minute on it, but at what cost?
3 trillion dollars, more money than you or I will ever have in our lifetimes, is predicted to be spent on AI data centres between now and 2029. More than 10,000 square feet, popping up like a bad game of whack-a-mole. All around the world. The only problem is, no one is there, paying attention to whack them down. According to IBM, an AI data centre is “a facility that houses the specific IT infrastructure needed to train, deploy and deliver AI applications and services. It has advanced compute, network and storage architectures and energy and cooling capabilities to handle AI workloads”.
Basically, every time you type a question into ChatGPT, or any other generative AI platform, the conversation that you’re having with the AI chatbot is able to occur because of the data centres that process what you’re asking and work to deliver a response. The cost of that response is electricity, specifically 10 times the amount of electricity it would take to ask Google the exact same question. So, when you consider that Chat GPT has 800 million weekly users, it’s unsurprising that a typical AI data center uses as much electricity as 100,000 households. It’s not just electricity though. Water is used during construction and once operational, to cool electrical components within the data centre. Researchers from the University of California say that AI-related infrastructure may soon consume six times more water than Denmark, a country of 6 million people. Don’t forget about the electronic waste! Hazardous substances like mercury and lead are runoff from these facilities, poisoning the environment.
It’s easy to just list off these impacts and forget about them, but for those who live in proximity to one of these massive facilities, there is no forgetting. In some parts of the United States, those who live around data centres built by Google, Meta and Chat GPT can’t even live in their homes because their water has been clogged by sediment. In Montreal, an Amazon data centre is refusing to explain how much water they’re using for cooling, and the city of Montreal couldn’t calculate a number due to how the building was constructed. Electricity bills are skyrocketing as well. The demand for electricity has increased so dramatically that those who reside around a data centre are paying for the company’s use because the billing is designed to be shared by those on the same grid. This crisis is getting closer and closer. In Etobicoke, Microsoft is currently building a data centre with a floor area of 280,000 square feet. It’s expected to use 39.75 litres of water per second for cooling purposes, according to planning documents submitted to the city.
While my lecturing may sound unnecessary and “extra”, I want all of us to be able to live out long, healthy and happy lives without worrying about something as crucial to life as water. We’ve been privileged for so long to have electricity, water and other natural resources that we’ve forgotten that it can be taken away from us. We're stuck in a game of whack a mole but the moles just keep popping up and no one is there to stop them. So we reach the question, how do we stop them? It's simple. 1 - limit our use of AI. Most of us are lucky enough to be able to attend school for free for 18 years of our lives. Applying the knowledge we learn in school allows us to actually get an education and retain information better. 2 - educate others. Even if only one person learns about the impact of AI and shares that information, we can reach a wider net of people than you might expect.
So next time you have an assignment due last minute, remember that using AI might win you the battle but it will definitely not win you the war. Grab that piece of paper, and use your own voice. Save our planet, one less prompt at a time.
References
https://www.ibm.com/think/topics/ai-data-center
https://www.npr.org/2025/10/14/nx-s1-5565147/google-ai-data-centers-growth-environment-electricity
https://www.youtube.com/watch?v=3Yh9OddmgS0
https://techcrunch.com/2025/10/06/sam-altman-says-chatgpt-has-hit-800m-weekly-active-users/
https://www.cbc.ca/news/ai-data-centre-canada-water-use-9.6939684
https://www.bbc.com/news/articles/cy8gy7lv448o
https://www.bloomberg.com/graphics/2025-ai-data-centers-electricity-prices/
