Working Notes is a separate track from the rest of this blog. Where the other series here, on green building, project-based climate resilience, and program design, are written for educational and organizational leaders evaluating what ILE Strategies offers, Working Notes is closer to a notebook: observations, half-formed arguments, and questions I’m still sitting with as someone who works at the intersection of climate, education, and AI. Expect fewer citations and more thinking out loud.

By Anasa Laude, LEED BD+C 

This piece follows up on themes from my book, AI with Purpose: A Practical Guide to Purposeful Artificial Intelligence for Social Impact.

I’m sitting in Brooklyn right now, deciding whether to finally buy an AC unit. It’s hot, but that’s not really the problem. The problem is that New York is currently blanketed in wildfire smoke from Canada, and our air quality index is sitting at 209, a number that would be alarming if I had any real sense of what it meant beyond “not this” (zero to fifty is considered normal). That’s the kind of moment that produces an essay, apparently.

Writing a book called AI with Purpose and then turning around to point out that AI might be quietly cooking the planet feels a little like naming your dog Gentle and then watching him eat the couch.

But here we are. Wildfires, floods, and other climate impacts require data and ways to process it, and AI has democratized access to that data. Yes, there are issues with AI fabricating information, biases in data, and its use in war and surveillance is worrying. Climate change itself may worsen because of lack of innovation in how data centers operate: the water use, air quality, and noise pollution involved, and the practices of tech companies and their allies in political seats who use eminent domain to displace homeowners in the footprint of planned data centers. These are problematic.[1]

When I studied to take the LEED BD+C exam in 2022, data centers featured prominently in the criteria. I learned a lot about the measures and criteria for sustainable data centers, but AI wasn’t top of mind then, and as I do continuing education to maintain the credential, I don’t see much innovation. LEED and USGBC’s own guidance measures a data center’s energy and water efficiency inside its own walls: how much power the servers draw, how much water the cooling system uses per square foot. It has no credit category for what happens outside those walls. It doesn’t touch water pollution from cooling discharge, the energy costs a utility passes on to ratepayers to build the transmission lines a data center needs, or the quality of life of the people who now live next to substations, transmission corridors, and the noise and traffic that come with them. A facility can be LEED certified and still be the reason a neighbor’s water bill doubled or their backyard became a utility corridor. As far as I know, not much has been done to close that gap.[10]

That said, AI (long before its debut for popular use) has been used in climate analytics and models now available to us all, and has made it possible for us to access data on precipitation, air quality, and water tables.[2] It allows us to translate this information into many languages so that people in remote parts of the world can access it and apply it.

In AI with Purpose, I mentioned Taina GPT, a local AI model co-developed with indigenous communities in the Amazon to create a repository of information on weather patterns, flora, and fauna. Tainá was built with Mura, Sateré-Mawé, and Tikuna communities in Manaus and has since grown into a second, community co-developed version.[3][4] Taina and other tools that have been developed to support research on climate trends and impacts have also aided emergency alert systems when flooding or drought is expected, allowing communities to plan. Recently in Venezuela, communities were able to use data and AI to pinpoint the most impacted areas and develop rescue efforts around them, with one existing app even repurposed on the fly into an early-warning system for future floods and landslides.[5][6]

It’s important that as we discuss the role of AI, the good, the bad, and the ugly, we keep in mind its uses domestically and globally. Yes, hold companies accountable, and enable decentralized AI owned and managed by local, small businesses, the kind of community-owned model where data and infrastructure stay under local control rather than a distant corporation’s.[7] For communities that welcome data centers (if any exist), explore shared ownership models. But climate change research, mitigation, and adaptation exist in this weird space in which cuts to science funding, access to U.S. data (or funding to gather and maintain that data) has been eliminated, the current administration having moved to eliminate NOAA’s climate research arm entirely and proposed zeroing out federal climate research funding in the 2026 budget.[8][9] And frontier AI companies are behaving badly, encroaching on communities with data centers and engaging in intense lobbying to ensure limited regulation, all of which threatens to exacerbate climate change itself.

I don’t have an answer to this, and I’m fairly sure nobody does, not even the very confident people on stage at tech conferences. But those of us concerned with the risks and opportunities of AI as well as climate change must weigh issues from all sides and work together toward common solutions, multilaterally and globally, with the most impacted communities at the center of these conversations. Consider this less a tidy conclusion and more me thinking out loud in front of you, which, if you have read this far, you have apparently agreed to sit through.

Sources

1. Rutledge, K. “The data center boom is transforming Georgia. Some residents could lose their land.” Atlanta Journal-Constitution, May 2026. https://www.ajc.com/business/2026/05/the-data-center-boom-is-transforming-georgia-some-residents-could-lose-their-land/

2. “Artificial Intelligence-Driven Analytics for Monitoring and Mitigating Climate Change Impacts.” MDPI, August 2025. https://www.mdpi.com/2673-4591/108/1/7

3. “Indigenous Groups Are Safeguarding Culture with Their Own ChatGPT.” Atmos, July 2025. https://atmos.earth/indigenous-groups-are-safeguarding-culture-with-their-own-chatgpt/

4. “What we learned by listening: deploying conservation AI in the Amazon.” GainForest, June 2026. https://gainforest.substack.com/p/what-we-learned-by-listening-deploying

5. “AI powers citizen-led earthquake response for Venezuela.” Rest of World, July 2026. https://restofworld.org/2026/venezuela-ai-citizen-disaster-response/

6. “Venezuela’s earthquake response was not built by the state or the UN.” Silicon Canals, July 2026. https://siliconcanals.com/sc-d-venezuelas-earthquake-response-was-not-built-by-the-state-or-the-un-it-was-assembled-in-three-hours-by-diaspora-coders-with-claude-and-replit-and-that-inversion-is-the-real-story/

7. “Community-Owned AI.” Sustainability Directory, November 2025. https://prism.sustainability-directory.com/area/community-owned-ai/resource/1/

8. “Trump seeks to end climate research at premier U.S. climate agency.” Science/AAAS. https://www.science.org/content/article/trump-seeks-end-climate-research-premier-u-s-climate-agency

9. “Proposed NOAA Budget Calls for $0 for Climate Research.” Eos, July 2025. https://eos.org/research-and-developments/proposed-noaa-budget-calls-for-0-for-climate-research

10. “Applying LEED to data center projects.” U.S. Green Building Council, updated May 2026. https://support.usgbc.org/hc/en-us/articles/12154267763987-Applying-LEED-to-data-center-projects

“What we learned by listening: deploying conservation AI in the Amazon.” Anasa Laude is a LEED AP BD+C credentialed educational consultant and co-founder of ILE Strategies, an NYC MWBE-certified consultancy focused on sustainability education, project-based learning, and thoughtful technology integration in K-12 and higher education. She is the author of AI with Purpose: A Practical Guide to Purposeful Artificial Intelligence for Social Impact.