If you have been following the growing coverage of AI and sustainability, you have probably noticed a familiar pattern. The conversation often lands on guilt and abstinence. People hear about energy-hungry data centers, water use, and emissions. The conversation about AI and sustainability often ends with a simple conclusion: maybe we should just stop using it.
That conclusion feels responsible, but it isn’t strategic.
AI has become foundational to how work gets done, how science advances, and how communities solve problems. Opting out in schools does not reduce society’s demand for AI. It simply leaves students less prepared to participate in a world in which AI is shaping civic life, careers, and problem-solving.
The real issue is not whether we use AI. The issue is whether the AI industry is required to build and power it responsibly.
We have been here before. We did not stop driving cars because of the pollution; we required emissions standards and redesigned engines. When refrigerants harmed the ozone layer, we did not abandon refrigerators, we phased out Chlorofluorocarbons (CFCs) and improved the technology. We did not stop eating packaged food because we lacked information--we required nutrition labels.
Nutrition labels do not require abstinence. These require transparency. Environmental labels can do the same for AI. Rather than opting out, we should demand accountability and redesign.
What Better Can Look Like
Some major companies have set clear sustainability targets. These examples give educators and students something concrete to point to.
- Google’s 24/7 carbon-free energy goal (by 2030)
- Google’s data center sustainability and efficiency work
- Microsoft’s sustainability commitments, including water positive and zero waste goals
- Microsoft’s carbon negative commitment (by 2030)
Examples of AI being used to reduce waste and improve efficiency within infrastructure itself are also available:
- Google DeepMind’s data center cooling work
- GraphCast, an AI model for faster and more accurate global weather forecasting
None of this erases the environmental impact of today’s AI, especially as demand accelerates. What it does show is that companies can be pushed to pair AI innovation with clean energy, water stewardship, and clear reporting.
Students and Staff Have Already Proven They Can Move Systems
Where I work, in New York City, student and staff driven environmental work is not theoretical. It has happened repeatedly with initiatives such as:
That same energy can be applied to sustainable AI.
A Strategy for Teaching AI and Sustainability
AI’s environmental impact is real, but product-level data is not surfaced in a simple, student-friendly way. That is what needs to change. We can start by building students’ understanding and then channel that learning into action: defining what transparency should look like.
These ready-made resources can ground students in the tradeoffs of AI and sustainability:
- SubjectToClimate: ELA lesson on generative AI’s energy use
- Share My Lesson: Mapping U.S. data centers and their environmental impact
- Scholastic Upfront: “Power Hungry” lesson plan
- National Science Teaching Association (NSTA): The carbon cost of our clicks
Once students understand the landscape, what could come next is to define what responsible disclosure should look like, and push the market in that direction.
Design an AI Environmental Label
Students already recognize standardized disclosures, such as nutrition labels. Privacy labels and indexes such as those from Common Sense Media, ISTE/ASCD, and Apple show how complex issues like student data practices can be made more transparent.
Pose a Challenge
If food and apps can be labeled, why shouldn’t AI tools have a label disclosing environmental impact?
Students can design a prototype AI Environmental Label that includes:
- Where the tool runs (cloud provider or infrastructure)
- Whether the company publishes sustainability reporting
- Whether there is a renewable or carbon-free energy target
- Whether water use is disclosed
- Date of last environmental update
- Third-party verification (yes or no)
Students can encourage rating platforms to include the environmental standard schools and communities should expect.
The Point
AI has real environmental costs today. Educators should not deny that, but opting out is not a strategy. Leverage is.
Schools and students can vote with their choices by favoring AI providers that publish credible sustainability data and commit to clean energy goals.
The next move should be as normal as nutrition labels: environmental ratings for technology tools. Until that exists, students can help create demand by proposing an AI Environmental Label and using it to push vendors toward cleaner infrastructure and clearer reporting.
This approach addresses AI’s growth responsibly while building students’ literacy, agency, and civic influence.
