There is no escape from generative AI. Even those who don’t actively use it (or don’t use it at all) will still hear of it and be impacted by it one way or another. From getting swamped on LinkedIn with variations of “If you are not doing this with AI, you’re already behind”, to your friends and family engaging in whatever image-slop AI trend is popular at the moment.
The rapidly growing use of generative AI has infiltrated both social and professional aspects of our lives. But efficiency is a hot buzzword for a reason, and with the continuous goal of corporations and people to get things done faster with less legwork, generative AI seems like a solution–if, of course, you are aware of the implications of the damage.
There’s a LOT to say about the impact of generative AI in our society. As a social impact and social justice marketing agency, we are always having conversations around the ethics of these tools.
But before we dive into our society approaching “Wall-E” levels of generative AI dependency, we thought for this piece we’d reframe the conversation to not talk about how it can serve us, but what happens when it does. We know money doesn’t grow on trees, we should also know AI doesn’t happen magically.
The Destructive Force Powering Generative AI
The force that powers generative AI is the same one that makes the internet happen: data centers. Data centers are warehouses that can sprawl over hundreds of thousands of square feet, housing rows upon rows of computers that store, process, and manage our data. They are big, they are hot, and they require an insane amount of energy and water to operate.
Big Hot Data Centers Are Thirsty for Our Resources
To understand how generative AI impacts our environment, we first need to understand how data centers work.
These facilities are designed to keep computers running at optimal performance, and for anyone who’s ever fired up The Sims on a computer, you know they tend to heat up. So not only do data centers need enough electricity to power all their technology, they also need water to cool it. And neither spawns out of thin air.
Generative AI Demands More and More Out of Data Centers
With the rise of generative AI, more data centers are being built to meet the growing demand.
The United States leads the global charts with over 5,000 data centers in 2025, followed by Germany in second place, with just over 500. Because we know there’s nothing America loves more than topping the charts over the most jarring aspects of our society!
And if that difference hasn’t shocked you enough, that’s almost a 3k increase from our data center numbers in 2023, which were around 2,500.
Although we can’t 100% attribute this growth to generative AI, we can safely assume that it played a major role due to the amount of power density it requires to operate.
According to Noman Bashir, a Computing and Climate Impact Fellow at MIT Climate and Sustainability Consortium (MCSC), “a generative AI training cluster might consume seven or eight times more energy than a typical computing workload”.
To put it in simple terms, generative AI is very hungry for power and needs its belly full to be able to work. To satisfy this endless hunger, more data centers are being built, leading to increased electricity and water consumption.
Soak It Up: A Single ChatGPT Query Eats up (est.) 5x More Electricity Than a Simple Web Search
A recent report from MIT estimates that the “power requirements of data centers in North America increased from 2,688 megawatts at the end of 2022 to 5,341 megawatts at the end of 2023… and by 2026, the electricity consumption could approach 1,050 terawatt-hours.”
That’s a lot of numbers, but when put in perspective, they are just as overwhelming. Just 1 terawatt-hour could power the entire state of California for 1.5 weeks. 1,050 terawatt-hours? That could power California for over 30 years.
According to the same report, “researchers have estimated that a ChatGPT query consumes about five times more electricity than a simple web search.”
And yes, the fact that generative AI is publicly available and people have started using it for everyday life definitely has an impact. But what really drives the power consumption is the amount of energy required to train these models.
Every single update and fine-tuning requires retraining the AI model. Think about how many updates ChatGPT releases a week, or how many organizations have started to build their own AIs. Think about how much energy we consume training these programs to learn how to simply summarize our emails better. How long could that power California for?
The amount of pressure put on our energy grids and water supplies to support the operations of data centers is incredibly high, and our communities and environments are starting to feel it.
Is Generative AI Worth the Detriment of Our Communities?
Just last year, a research paper by Harvard and the University of Pisa reported that “U.S. data centers produced 105 million tons of CO2 equivalent gases in the past year with a carbon intensity 48 percent higher than the national average.”
Another Energy Department report noted that in 2023, the total water use by data centers had tripled since 2014, to 66 billion liters (or 17.4 billion gallons).
Every day, communities suffer from the continuous depletion of resources. Local utilities are becoming more expensive due to high consumption. The decrease in air quality is affecting the health of those living near data centers. Cultural areas are diminishing so that corporations can make space for the data centers needed to power this insatiable tech monster.
If you think this hasn’t affected you, or that it won’t, reconsider. Considerable damage has already been done already and there are no regulations in sight to mitigate this harm.
So all of this begs the question: how much further does all this need to go before the masses start caring? Is the strive towards efficiency really worth this environmental strain?
As we continue navigating a world where generative AI is common usage, we encourage questioning your generative AI usage and reflecting on its impact.
Change starts with us. And if we want a world where we can live comfortably and safely, we could all use more humanity and less generative AI in our lives.
If you are curious about how you can start limiting your usage of ChatGPT and generative AI, we put together a post with our top 5 Alternatives for many common use cases.
