How much should teachers use AI? A few questions to consider…
I remember a few years ago when ChatGPT first showed up—schools were literally losing their minds, first, trying to prevent students from using it altogether and then months later, trying to detect when they did and soon after, to what extent. The programs we used scanned a student’s writing and identified what percentage, if any, was generated by AI. I’m not going to lie, these programs seemed effective—for a few months anyway—until new AI programs enabled students to use AI to disguise their previously generated AI. As you can imagine, it was extremely messy and unnerving, and there were precious few guidelines for schools and teachers.
Nonetheless, at that time, I believed that AI’s potential to move us forward in our thinking and in our writing outweighed its downsides. It seemed to me to be a more sophisticated Google and I immediately thought of it as a way to get up to speed on a topic and begin the thinking and writing process at a more advanced place. Why wouldn’t we want to start any writing or research project with a brief and seemingly accurate synopsis of what has already been written about that topic? As for my students, I immediately came up with some guidelines for use in my 12th grade English classroom and urged my students to try it for brainstorming and for getting background on topics they were looking into. In fact, almost exactly three year ago, I wrote about it here: https://www.teachingandbeing.com/blog/
A lot has changed since then.
In many high school classrooms, including mine, some students use AI for everything, sometimes unwittingly, often stealthily, and I have come to believe that the only meaningful way to help them to become better readers, writers and thinkers is to to engage them deeply—away from their computers—and, more importantly, away from AI.
What this means is that I spend a lot of class discussion time convincing them there there is real value in being effective communicators; that their vocabularies—the words they use regularly—greatly impact how the world sees them; that being a good close reader and listener will help them to be successful in almost any field; and that understanding and referring to other people’s ideas in their writing and conversation is a critical skill for college—and for life. By shifting the focus from the finished product to the complicated, painstaking process of thinking and writing, I try to help my students to understand that while AI might seem like the quickest most reliable answer, it circumvents many cognitive steps and in doing so, actually degrades their ability to reason and to think critically about how they got to the answer. Quite a bit of neuroscience research underscores this idea, with a broad consensus emerging that an overreliance on AI can lead to: decreased creativity, shorter attention spans. poor critical thinking and limited retention. (https://www.bbc.com/future).
With all of this in mind, I remind myself that not only are my students learning what I teach them, but they also look to me to model effective literacy strategies. I do this in the way I speak and write to them, the examples I share with them, the assignments I create for them and in the ways that I evaluate them and respond to their work.
And this really leads me to my first big question.
Are teachers aligning their use of AI with their most essential education values?
If teachers are advocating for authentic learning and deep engagement with texts and ideas, we need to walk the walk. While many of the AI tools that I see English teachers using right now promise to help teachers save time and become more efficient, very few promise to help students delve deeper. Study after study suggests that the best way to improve student writing is with individual conferencing and personal comments that speak to the writer, not the writing. Why, then, do I want to rely on an AI program that evaluates student writing and generates comments, most of which we know students do not read and do not help them to become better writers.
Granted, the powers-that-be do have a necessary interest in making sure that instruction is reasonably standard and that education outcomes are meaningful and quantifiable, but the truth is the best learning is not always quantifiable. AI can be most useful, I think, in analyzing broad swaths of data but like other standardized metrics, it has significant limitations in the day to day business of helping individual students to learn. Many educators I follow share this view, but in many schools, including the one I work in, the conversations about AI are focused almost entirely on the possibilities of AI—and the ways it can improve efficiency. As a result, many teachers are jumping on board with these programs without really interrogating how they align with the most important values of our profession.
Are teachers being transparent enough about their use of AI?
And guess what?! Our students know. It turns out that even without a detection program, it’s become fairly easy to recognize AI-generated writing, whether it’s in a student essay, an email or assignment guidelines. But the worst thing is, we’re not talking about our use of AI with our students. In a 2024 survey (EducationWeek) 80% of teachers who used AI regularly said it was not necessary to disclose their use of AI to students or administrators when planning lessons and 48 percent said this was the case even when using AI for grading and feedback. My question is, if we are not being open and transparent about our use of AI, are we being ethical? And, what message does this send to our students about the ethical use of AI?
Are teachers really considering the human and environmental costs of AI?
My final question is the thing that got me thinking about this topic in the first place: The terrible damage that the production of AI is doing not only to our planet, but also to the people who live in communities where data centers are being built. To this point, the National Education Association (NEA) released a statement last week, urging teachers to consider the environmental impact of AI’s use and to adopt sustainability practices:
Frequent AI-powered tasks—such as automated grading, image generation, adaptive learning, and chat bot interactions—consume considerable energy and contribute to the need for more data centers in communities.
You should be aware of the carbon footprint associated with AI tools and advocate for sustainable options.
Where possible, schools and universities should adopt policies that prioritize energy-efficient AI models and cloud technologies powered by renewable energy.
You should teach your students about the environmental impact of AI as part of learning around digital literacy.
When you discuss AI ethics with students, you should include sustainability and responsible AI usage.
When you create assignments and projects, you should encourage your students to explore energy-efficient AI alternatives, when possible.
AI should complement, not replace, traditional teaching methods. Hybrid approaches that combine AI-driven personalization with resource-efficient teaching will help mitigate environmental costs.
Finally, I will close with what I think we can all agree is the most profound cost of the production of AI: the human cost. As with many environmental assaults, this one lands squarely on the lives of poor and underrepresented communities. A recent National Institute of Health report (https://pmc.ncbi.nlm.nih.gov/articles) concluded that the global expansion of AI has been guided entirely by surging demand, rather than by sound policy to protect public health. According to this report, there is an urgent need for empirical data to understand long-term health effects, but this is what we know so far:
Data centers generate untenable noise pollution which adversely affects workers, community members and wildlife;
Air pollution is the most acute concern. Fossil-fueled power plants and diesel backup generators that power data centers emit hazardous pollutants such as nitrogen oxides and fine particulate matter, increasing rates of respiratory diseases, cardiovascular conditions, and elevating cancer risk in nearby communities. A recent model indicates that the U.S. data centers in 2030 could contribute to nearly 1300 deaths annually, resulting in a public health burden of more than $20 billion.
Data centers rely on often-scarce local water supplies, which exacerbates water insecurity and increases the risk of waterborne diseases, dehydration, and poor hygiene in affected communities.
To be clear, I am not advocating for a complete abandonment of AI’s use in schools and classrooms, but I am suggesting that we teachers do our due diligence in understanding not only the possibilities of these emerging technologies, but also the profound costs.
Further Reading
https://www.nea.org/professional-excellence/student-engagement/tools-tips/environmental-impact-ai
https://sites.uab.edu/humanrights/2025/10/02/
https://www.sciencedirect.com/science/article/pii/S2950138525000178
https://newsroom.ucla.edu/stories/opinion-ai-is-destroying-our-planet-we-must-act
https://www.bbc.com/future/article/20260505-how-to-use-ai-without-turning-your-brain-to-mush
https://www.edutopia.org/article/teachers-use-ai-3-troubling-patterns/
