Howdy, post-graduation! I had the official RAND graduation last week which was oddly emotional. After 9 months, it’s hard to believe the second Master’s degree of Technology Policy is over. I made a ton of friends along the way and I’ll still be here in DC. Who knows? Maybe, they might gimme a call if they need a lecturer about data centers, environmental impacts, and generative AI? 😄 I am sorta an expert after all!

A lot happened this past week in the news and not all of it was good… Turns out the DOJ finally weighed in on the public health impacts court case with xAI, the U.S. may be letting a key data center regulation expire in September, Europe has to choose between its climate goals and AI, and FERC finally issued some guidance on data centers to grid operators.

Chau 👋,

Nate

The Primary Energy Fallacy: Why Electrification Changes Everything, EP. 293 (AI & Energy): A deep dive into how electrification is reshaping transport, buildings, and industry, why electric technologies are often far more efficient than their fossil-fuel counterparts, and what that means for energy demand and system planning over the coming decades.

Will AI Help or Hurt Climate Decision Making | Michal Nachmany | TEDxLisboa (AI & Climate Science): After two decades working on climate data and tools, Dr. Michal Nachmany asks whether general-purpose AI can be trusted to surface everything relevant, admit when it doesn't know, and be honest about the choices it makes on our behalf. Her takeaway: what matters is how these tools are designed, and who they're designed for.

The Week in Green Software: State of The Fossil-Free Internet (AI & Environment): Kate and Tzviya dig into the Green Web Foundation's State of the Fossil-Free Internet 2026 report, covering the limits of carbon accounting, the rapid growth of AI data centers, and why better measurement and clearer reporting are essential for a more sustainable internet.

Have Data Centers Raised Your Electric Bill? Causal Evidence from the United States (AI & Energy): Using an instrumental-variables approach, the authors find data centers actually caused average U.S. retail electricity rates to fall modestly from 2015 to 2024, as durable demand growth spreads fixed costs and unlocks economies of scale in transmission, distribution, and generation. They caution that future supply constraints could reverse the effect. Authors: Watten et al.

Net Environmental and Economic Impacts of AI-Powered Microgrids: When Context Matters More Than Technology (AI & Energy): Drawing on two years of operational data from eleven microgrids across Europe, North America, and Australia, the authors find AI optimization cut energy costs by 18.3% on average but delivered net environmental benefits in only three of eleven cases, with outcomes ranging from -290 to +1,400 tCO2e/year. Authors: Paccou et al.

Green Artificial Intelligence: A Comprehensive Review of Metrics, Tools, Challenges, Trends, and Future Prospects (AI & Data Centers/Hardware/Model): A wide-ranging review of Green AI, examining low-power architectures, model compression, specialized hardware like TPUs and spiking neural networks, and edge computing as routes to lower energy costs and higher efficiency. Authors: Peykani et al. Journal: Archives of Computational Methods in Engineering.

Data Center Construction Creates a Resource Shock (AI & Infrastructure): A Bank of America Institute analysis of how the data center construction boom is straining power, water, and critical materials, and where the bottlenecks and investment opportunities lie. Authoring Body: Bank of America Institute.

Ratepayer Protection Act (AI & Energy): Bipartisan legislation that would amend the Public Utility Regulatory Policies Act of 1978 to require large electricity users like data centers to cover the full cost of the grid infrastructure needed to serve them, shielding households from the bill. Authoring Body: House Committee on Energy and Commerce.

United States Data Center Energy Usage Report: 2025 Update (AI & Energy): Lawrence Berkeley National Laboratory's updated accounting of how much electricity U.S. data centers actually consume, and how fast that figure is climbing. Authoring Body: Lawrence Berkeley National Laboratory

Data Centers Get the Headlines While Life Sciences are Quietly Booming | Spring 2026 (AI & Multiple Applications): A webinar on what's really driving the life sciences market right now, from economic headwinds and shifting supply chains to surging manufacturing activity and the ripple effects of competition from other sectors.

I Tracked Down the Hidden Workers Secretly Powering ChatGPT (AI & Ethics & Misinformation): An investigation into the hidden workforce training the AI that may replace them, where nearly 1 in 5 workers have fallen into homelessness, exposing a new gig economy run by Big Tech.

AI and the Energy Transition: Implications for U.S. Economic Growth and Inequality (AI & Energy): Costa Samaras of Carnegie Mellon presents on the interconnected risks and opportunities of AI expansion and the energy transition, followed by a panel on how policymakers should navigate rapid economic transformation.

Spore.Bio (AI & Environment): Building a new generation of microbiological technologies with AI. Location: Europe.

BioEsol (AI & Energy): Innovative battery solutions for businesses transitioning to renewable energy, built for a reliable, optimized, and cost-effective supply. Location: South America.

Telaqua (AI & Water Resources): The AI infrastructure for water operations. Location: Europe.

ClimateQ&A MCP (AI & Sustainability): An MCP-server version of Ekimetrics' ClimateQ&A, an LLM assistant that answers climate and biodiversity questions grounded in IPCC and IPBES reports, ready to plug into AI assistants and agent workflows. Price: FREE.

Brazos (AI & Infrastructure): Google's open-source liquid-to-air cooling sidecar that brings high-performance liquid cooling to existing air-cooled data halls one rack at a time, supporting a 60 kW thermal load per rack without re-plumbing the building. Price: FREE.

PNNL/ASHRAE/NEMA AI Data Center Energy Performance Framework (AI & Infrastructure): A collaborative framework offering scalable, non-mandatory guiding principles for the efficient design, commissioning, retrofit, and operation of AI data centers across their full lifecycle. Price: FREE.

If you have resources that you wish to submit to the AI & Environment Resource Hub, fill out the Google Form! Thank you for suggesting resources to add to the Resource Hub. 🙃

For any anime fans out there, this Re-Zero/data center crossover meme is literally Chef’s kiss 🤌🤌🤌🤌🤌🤌🤌🤌🤌🤌🤌🤌

Thanks for reading the 46th issue of The Climate Code! It means a lot to me and we have more coming in the future, so definitely stick around! 💚

Nate

P.S.

New here? Check out the AI & Environment Resource Hub if you have no idea what it is. 😆