Exploring the Hidden Environmental Cost of Generative AI

Generative AI can be an incredibly useful educational tool. From prompting creative thinking to summarising academic texts, AI tools have changed the way that students experience learning inside and outside the classroom. Generative AI is still so new that the full extent of its capability is still being explored. Already, there is a great amount of discourse surrounding the usefulness and potential of AI compared to the possible danger of it becoming a substitute for critical thinking. Aside from these concerns, there is an unseen danger to the power of generative AI - its environmental footprint. 

Where Does the Power Come From?

It can be difficult to understand how technology impacts the environment without seeing the chains that link an action to its environmental consequences. It’s easy to see how the devices we use can impact the environment, since there is no easy way to recycle them. On the other hand, it can be tricky to see how asking a virtual chat bot to generate a picture of a unicorn flying through space could impact the environment, since the results are so instantaneous. We don’t tend to think about where the power to create that image comes from. 

How Much Energy Does AI Generation Require?

The computational power needed to train AI models and generate images and text can require huge amounts of electricity. Most of this electricity is generated by burning fossil fuels. This increases CO2 emissions, puts strain on the electrical grid, and uses tonnes of water to keep the machines cool. A study done by Hugging Face and Carnegie Mellon University revealed that generating a single AI image could use the same amount of energy as half a smartphone battery charge. While this may not seem like much, estimates put the number of AI images generated per day at 34 million. Since AI images take only seconds to generate, the amount of energy used compared to the results is staggering. 

How Should We Move Forward?

As generative AI becomes a bigger part of our everyday lives, from education to entertainment, it’s important that we start actually considering the hidden costs of convenience. Just as we teach students to consider the sources of their information for reliability, we should also encourage awareness of the environmental cost behind the tools they use.

This doesn’t mean we need to stop using AI altogether but we should approach it responsibly. That might involve using AI more mindfully, supporting companies that invest in greener infrastructure, or simply starting conversations about the sustainability of digital tools. If we want to build a future where technology supports both learning and the planet, we need to keep talking about it. Awareness is the first step toward making smarter, more sustainable choices.