Weekly News Wrap #2
Welcome to your weekly briefing! This past week has seen significant discussions and developments at the intersection of digital innovation and the circular economy. Our top stories highlight the growing energy demands of AI infrastructure and the urgent need for more physically matched renewable energy solutions, as well as exciting advancements in digital tools for sustainability reporting and circular materials innovation in product design. From real estate to apparel, technology is increasingly proving to be a critical enabler (and sometimes, a challenge) on the path to circularity. Listen to our AI-powered podcast here.
Here’s a roundup of the most relevant news:
Weekly Roundup & Key Insights
- AI’s Energy Footprint: The booming demand for Artificial Intelligence is driving a surge in energy consumption, particularly from data centers, threatening global emission reduction efforts. New research forecasts that the leading AI supercomputer in 2030 could require as much as 9 gigawatts (GW) of power, equivalent to nine nuclear reactors. This highlights a crucial challenge for the digital economy to become more sustainable.
- Efficiency Gains are Crucial: While the overall power demand is rising, there are innovations. AI supercomputers are becoming more energy efficient, with computational performance per watt increasing by 1.34x annually due to more efficient chips. Data center operators are exploring liquid cooling and AI tools for real-time temperature optimization to boost efficiency.
- Digital Tools Accelerate Circularity Measurement and Reporting: We’re seeing rapid adoption of software solutions designed to measure, manage, and report on sustainability. Measurabl’s free global sustainability software solution for real estate, for instance, has been adopted by 11,000 buildings in just four weeks, enabling tracking of energy, water, waste, and carbon. Nasdaq Metrio™ is also cutting sustainability report creation time in half through data auto-population. Carbonfact precisely measures the potential environmental impact of material substitutions and other decisions, aiding their decarbonization agenda.The Carbon Crediting Data Framework is an open-source initiative designed to bring structure to the voluntary carbon market data. For circular startups seeking funding, DEFINITE-CCRI offers a free self-evaluation tool to assess investment readiness and simplify the funding process.
- Circular Materials in Tech & Textiles: Acer has launched the Aspire Vero 16 laptop, featuring a chassis made from oyster shells and over 70% post-consumer recycled plastic, marking a significant step in making technology itself more circular. In textiles, startups like Circ, Ambercycle, Reju, Samsara Eco, and Syre are pioneering textile-to-textile polyester recycling with goals of “infinite recycling”.
Spotlight report: AI's Energy Footprint & Sustainable Digital Infrastructure
The massive growth in Artificial Intelligence (AI) and cloud computing is leading to increased construction of power-hungry datacentres. These facilities are projected to increase energy demand by as much as 165 percent by 2030, straining existing electricity grids and potentially undermining broader climate action plans. Many of these new datacentres are relying on gas-powered facilities, with over 85 such projects globally to meet the burgeoning energy demands of AI. The International Energy Agency expects greenhouse gases from datacentre energy consumption to double in the next five years as a proportion of global fossil fuel emissions.
Despite claims by major tech companies to be “100 percent matched” with renewable energy, this often refers to financial investments in green power credits rather than a direct physical supply of clean energy to their datacentres. This means that while companies invest in renewables elsewhere, their datacentres can still draw power from local grids heavily reliant on fossil fuels, with the most polluting forms of energy often used to meet peak demand.
To address this, there’s a growing focus on boosting datacentre efficiency. Innovations include:
Liquid cooling systems, which are more economical than air cooling for advanced AI chips that generate significant heat. Amazon is implementing custom-made liquid cooling systems.
AI tools that tweak temperatures in real-time to optimize cooling and overall energy use in large data centers.
Advances in chip and power management hardware from companies like Nvidia, Groq, and AmberSemi, which aim to cut energy waste. AmberSemi, for example, uses “vertical power delivery” to conserve energy during power transfer to motherboards.
The concept of “large load flexibility”, where data centers reduce power usage by switching to on-site sources or deprioritizing non-urgent tasks during peak demand. Google is reportedly experimenting with this by varying energy use at its data centers in agreements with utilities. However, private companies currently have little incentive for this complex and potentially expensive load-shifting.
A crucial need identified is the reform of energy accounting to better reflect the physical reality of electricity grids, pushing datacentre operators to invest in and use renewable energy from local sources that match the timing of local energy usage. Google has expressed support for hourly matching of energy investments to energy use and has invested in CO₂-based battery storage that can dispatch clean energy for up to 24 hours.
The rapid growth in AI supercomputers means their power requirements and hardware costs are doubling every year, with industry ownership of these systems surging from 40% in 2019 to 80% in 2025. The United States hosts about 75% of global AI supercomputer performance. The immense power demands might eventually necessitate decentralized training approaches, distributing computing tasks across multiple locations. (Sources: FT, Trellis and EpochAI)
Innovations in Circular Production & Design
In the apparel industry, a new focus on circularity and safer materials is emerging. Companies like Carter’s, Gap Inc., and H&M Group are implementing various strategies. Carter’s has relaunched its KidsCycle takeback program, while Gap Inc. has partnered with textile recyclers such as Ambercycle and Syre and has kept nearly 850,000 items in circulation through its resale partner ThredUp. H&M Group is also pushing circularity through resale and investments in recycling startups. These efforts highlight the growing role of digital platforms for resale and advanced recycling technologies in closing material loops. Startups like Circ, Ambercycle, Reju, Samsara Eco, and Syre are betting on textile-to-textile recycling to create circular alternatives to virgin synthetics, with Samsara Eco aiming for “infinite recycling”.
Policy & Frameworks for Circularity
The UN plastics treaty negotiations are in their final round, aiming for a legally binding global treaty to limit plastic pollution across its entire life cycle, from production to waste. Key implications for businesses include new packaging and material restrictions, stricter rules on toxic chemicals, and increased demand for circular design and reuse models. This highlights the increasing regulatory pressure that will drive the adoption of circular economy principles, often requiring digital solutions for compliance and implementation.
The Circularity Gap Report:Finance 2025 indicates that the world is currently only 6.9% circular, emphasizing the urgent need for more resource regeneration and effective circular solutions beyond just recycling. Investment in the circular economy is growing, increasing from $10 billion in 2018 to $28 billion in 2023, though it peaked at $42 billion in 2021. While this investment represents only a small portion of total commercial investment (2%), the report’s numbers are likely an underestimate due to the difficulty in defining eligible financial instruments and the exclusion of bonds, which account for a significant portion of investment (15-25%). The analysis also highlights that investment is primarily downstream, and a need exists for public-private partnerships and addressing challenges like the ‘valley of death’ for early-stage ventures
