Happy New Year and welcome to the first Jisc Digital Sustainability Newsletter of 2025! Each month we delve into the latest news, trends, articles and insights at the intersection of technology and environmental sustainability.
If you haven’t already, be sure to subscribe to our DIGITAL-SUSTAINABILITY JiscMail mailing list to receive future editions of the Jisc Digital Sustainability Newsletter direct to your inbox.
News
Let’s kick things off with a roundup of some of the most interesting global digital sustainability news headlines and developments from the last month. Click on the headline to learn more:
UK Government Introduces Regulations to Hold Online Marketplaces and Vape Producers Accountable for E-Waste Recycling
The UK government has announced new regulations requiring online marketplaces and vape producers to contribute to the recycling and treatment of waste electrical items, such as vapes and household electronics. Previously, UK-based businesses bore the majority of recycling costs, creating an uneven playing field.
Meta to Build Largest-Ever Data Centre Powered by Fossil Fuels, Sparking Environmental Concerns
Meta has announced a $10 billion, 4-million-square-foot data centre in Louisiana, powered primarily by natural gas. Critics have raised concerns about the greenhouse gas emissions from the plants and the unproven feasibility of future hydrogen co-firing upgrades. While Meta claims the project will add enough renewable energy to the grid to offset the centre’s energy use, sceptics highlight the lack of comprehensive evaluations of cleaner alternatives, like wind power.
IET Report Highlights Public Unawareness of AI Energy and Environmental Costs
A study by the Institution of Engineering and Technology reveals that only 16% of the UK public are aware of the environmental impact of large language models like ChatGPT. The IET recommends that LLMs be used for complex tasks while simpler queries rely on basic search engines to reduce emissions. The report also calls for government-backed green data centre regulations and sustainable AI training methods. Encouragingly, 60% of respondents said they would adjust their behaviour after learning about AI’s environmental footprint.
Nokia Commits to Net-Zero Emissions by 2040 with Innovative Sustainability Strategies
Nokia has pledged to achieve net-zero emissions by 2040 while reducing its environmental impact and leveraging technology for societal benefits. The company has already cut its greenhouse gas emissions by 11% across Scopes 1, 2, and 3 and improved energy efficiency in its products, such as reducing power consumption of its 5G AirScale Base Transceiver Stations by 50%.
TikTok’s Carbon Footprint Surpasses Greece’s, Study Reveals
A study by Greenly reveals TikTok’s 2023 carbon emissions in the US, UK, and France totalled 7.6 million metric tonnes of tCO₂e, exceeding Greece’s total annual emissions. Despite its carbon-neutral goal by 2030 and a renewable data centre in Norway, TikTok has yet to match reporting transparency of peers like Meta and Google. Critics emphasise the platform’s limited progress on renewable energy adoption.
Harvard AI Cluster Named a Top Green and Fast Supercomputer
Harvard’s Kempner AI cluster has been ranked 32nd on the Green500 list for energy efficiency and 85th on the TOP500 list of fastest supercomputers. Housed at the Massachusetts Green High Performance Computing Centre, the cluster uses 528 GPUs and advanced parallel processing to support breakthroughs in AI and neuroscience, powered by 100% carbon-free energy from hydroelectric and solar sources.
Microsoft Shareholders Reject Proposal to Limit AI Use in Fossil Fuel Industry
Microsoft shareholders voted down a resolution by environmental nonprofit As You Sow that called for the company to assess risks associated with deploying AI technologies to support fossil fuel projects. Critics argue Microsoft’s AI and Energy Principles inadequately address environmental harms and fail to evaluate fossil fuel companies’ net-zero claims.
AI Data Centres May Face Power Shortages by 2025
Global research and advisory firm Gartner predicts that by 2025, 40% of AI data centres will lack sufficient power to operate fully without additional energy sources. AI energy demands could soar to 500 terawatt-hours annually by 2027 – more than double current levels.
E-Waste Crisis Endangers Lives in Ghana
BBC News reports that workers at refuse sites in Ghana are burning electronics to extract metals, releasing toxic fumes that harm health and the environment. The crisis underscores the need for stricter global regulations and sustainable solutions to protect vulnerable communities and reduce environmental harm.
Digital Sustainability at Jisc:
Now for a quick rundown of some of the digital sustainability initiatives we have been working on here at Jisc to support our members across tertiary education:
Driving Digital Sustainability Workshops – Places Available
February will see the launch of our first Driving Digital Sustainability workshop – a practical and engaging session designed for those wanting to learn more about what they as individuals and organisations can do to better understand and embed sustainability best practices across their digital activities.
This 2-hour online workshop is designed to go beyond theoretical discussions and instead offer real-world solutions through identifying high-impact low-cost initiatives. It provides a tailored action plan, practical tools, and clear first steps to embed sustainability into digital operations, helping organisations reduce their environmental footprint and uncover cost-saving opportunities.
Spaces are limited to 20 participants per session – so don’t miss out! You can find out more about the Driving Digital Sustainability Workshop here.
FE & Skills Digital Sustainability Community Meetup #4
Join us for the next bi-monthly FE & Skills Digital Sustainability Community Meetup on Tuesday 25th February at 2pm. We’ll be diving into two important topics:
- The Environmental Impact of AI: Catherine Barker, from Jisc’s AI Team will be helping us better understand the complexities of AI’s impact on the environment.
- Reduce Cloud Storage: An exploration of how we can better manage the way we manage our cloud storage.
If you’re not already part of the FE & Skills Digital Sustainability Community, it’s easy to join! Membership is open to anyone working in the FE or Skills sector – just fill out the community registration form to gain access to this and future events.
Resource Corner :
Each month, we share a digital sustainability resource that we hope you might find useful.
This month, we recommend AI and Climate: Balancing Efficiency and Emissions, a thought-provoking piece by S&P Global Sustainable1. This article delves into the dual impact of AI on energy use: its potential to drive energy efficiency and reduce emissions in industries like manufacturing, and its increasing demand on energy-intensive data centres. While advancements in renewable energy adoption by major operators are promising, the persistent reliance on fossil fuels raises concerns about achieving net-zero goals. The piece highlights the importance of balancing AI-driven efficiency gains with strategies to mitigate its growing environmental footprint.
You can explore the full report here.
Member Highlights:
In our monthly member highlights, we turn the green spotlight onto one of Jisc’s members in FE and HE who are implementing noteworthy digital sustainability initiatives.
Nottingham Trent University has committed to reaching Net Zero by 2040, including its supply chains. To meet this ambitious goal, NTU has developed the Net Zero Carbon Supplier Tool to support the university sector in reducing supply chain carbon emissions.
The tool not only helps suppliers capture and report their emissions data but also provides them with a free carbon reduction plan and resources to transition toward Net Zero. Regardless of company size or progress on sustainability, the intuitive tool offers tailored guidance to simplify the process. Suppliers can input their emissions data, upload reduction plans, and receive ongoing support through the tool’s annual reporting framework.
For more information, including a video tutorial on how the tool works, click here.
Digital Sustainability Articles:
Here are a selection of our favourite articles on digital sustainability from the last month. Click on the title link to be redirected to the full article:
Demand Shifting and Shaping in Green Software
Charles Humble explores two key techniques in green software: ‘demand shifting’ and ‘demand shaping.’ Demand Shifting moves tasks to times or regions with lower carbon intensity, optimising cloud usage by utilising cleaner energy sources. ‘Demand shaping’ adapts system behaviour based on current carbon intensity, such as reducing video quality during high carbon periods. Both methods help reduce energy consumption and carbon emissions while maintaining operational efficiency.
The Future of AI: Neural Networks in Hardware
ChandraKumar R Pillai delves into the potential of logic-gate neural networks, an innovative approach that embeds AI directly into hardware, offering a promising solution to AI’s energy consumption challenges. Traditional AI systems rely on software simulations, which are energy-intensive. However, logic-gate networks, by using hardware logic gates, reduce energy consumption significantly. While there are challenges, including training complexity and scalability, the shift to hardware-based AI systems could revolutionise efficiency and sustainability.
How to Create a Green Software Culture at Your Company
Annie Freeman highlights practical steps for tech companies aiming to reduce their environmental impact through sustainable software practices. To embed sustainability into workflows, Annie suggests a number of insightful and practical initiatives that can be read in the full article, such as building dashboards for emissions monitoring and optimising workloads with tools like Kubernetes.
ARC-AGI Triumph and the Eye-Watering Energy Bill to Match
Rabih Bashroush explores OpenAI’s latest breakthrough with the O3 AI model, which set new records on the ARC-AGI benchmark. While the performance is groundbreaking, Bashroush highlights a critical concern: the model’s staggering energy consumption. In its most intense run, O3 consumed enough electricity to power a UK household for over four years. As AI continues to scale, the article emphasises the urgent need to balance technological advancements with sustainable practices to mitigate AI’s environmental impact.
Women in AI: Tamar Eilam is helping IBM build sustainable computing
In an insightful interview with Dominic-Madori Davis of TechCrunch, Tamar Eilam, IBM Fellow and Chief Scientist for Sustainable Computing, focuses on reducing AI’s energy consumption. She leads projects like Kepler, which quantifies energy use in containerised applications. Eilam advocates for efficient AI model development and transparency in energy use.
Why We Don’t Know AI’s True Water Footprint
Miranda Gabbott explores in TechPolicy.press the hidden water usage of AI-driven data centres, with a focus on Spain’s rapidly expanding infrastructure. She highlights how cooling methods for these centres – essential for preventing server overheating – require significant water or energy, often at the expense of local resources. Gabbott calls for independent assessments, better planning, and a revaluation of AI’s role in consumer applications to ensure sustainable practices.
The Tiny Programming Tweaks That Could Slash Emissions
Chris Baraniuk, writing for The Reengineer, highlights how small programming adjustments can significantly reduce software energy consumption and carbon emissions. For example, Danny van Kooten’s removal of 20 kilobytes of code from a WordPress plugin reduced global emissions equivalent to cancelling 1,200 annual transatlantic flights. Similarly, optimising web content delivery and image sizes has cut website emissions by over 80% in some cases.
Will AI Save the Planet? Why the Evidence Is Flawed
Jo Lindsay Walton, writing for The Conversation, critiques optimistic claims about AI’s potential to tackle climate change. While AI is touted for its ability to model wildfires, optimise energy, and develop low-carbon materials, Walton highlights issues with overstated benefits and flawed research methods and discusses the challenges in comparing AI’s environmental costs to human activity, questioning the reliability of studies claiming AI’s sustainability benefits.
Why Sustainable Data Storage is the Future of IT
Jon Howes, Senior Vice President at Wasabi, writes in The Compute, Storage & Networking Channel about the critical role sustainability plays in data storage decision-making, particularly as regulatory pressures intensify. He highlights the mounting challenge for IT leaders: balancing increased reliance on data and AI with corporate sustainability commitments.
What Do You Lose When You Abandon the Cloud?
Cloud repatriation is gaining attention, with some companies moving workloads off public clouds to reduce costs. However, as Charles Humble discusses, this shift often overlooks critical trade-offs. Cloud providers enable scalability and innovation, but when moving back to data centres, firms face high upfront costs and difficulty in managing growth. The cloud also offers flexibility, optimising resource use on-demand, unlike on-prem solutions that require heavy forecasting and capacity planning. Humble suggests that while repatriation can reduce costs at scale, it might also hinder innovation and complicate security and sustainability.
Microsoft Introduces Zero-Water Cooling for Data Centres
Steve Solomon, Vice President of Datacenter Infrastructure Engineering, discusses Microsoft’s innovative zero-water cooling system, launched in August 2024. This new design, highlighted on the Microsoft Cloud Blog, recycles water through a closed-loop system to cool data centres without evaporating fresh water, potentially saving over 125 million litres annually per site. Although the new design increases energy usage compared to traditional methods, it leverages high-efficiency cooling technologies to mitigate power demands.
Why Sasha Luccioni Wants to Create Energy Star Ratings for AI
Issie Lapowsky, writing for Fast Company, offers a profile of Sasha Luccioni, Hugging Face’s climate and AI lead who is on a mission to raise awareness about the environmental impact of large language models. As part of the AI 20 series, this article highlights Luccioni’s unique role in bridging the gap between AI development and sustainability.
Podcast Pick:
This month, we’re introducing Podcast Pick – where, as the name suggests, we highlight a podcast episode that’s caught our attention.
Solving for Climate is a fantastic podcast by the incredibly knowledgeable Hannah Ritchie and Rob Stewart. This month we would love to recommend the episode, The Climate Cost of AI, in which Dr. Sasha Luccioni, AI and Climate Lead at Hugging Face, dives into the hidden carbon impact of generative AI tools and how we can maximise their benefits while minimising their environmental costs. Packed with insight and solutions, it’s a genuinely thought-provoking listen.
Get Involved:
We want to hear from you! Share your comments, suggestions, and digital sustainability highlights. Contact our Subject Specialist for Sustainability, Cal Innes, at cal.innes@jisc.ac.uk.
And don’t forget to subscribe to our DIGITAL-SUSTAINABILITY JiscMail mailing list for future editions of the Digital Sustainability Newsletter.