This situation has concluded
It was preserved as a record on October 27; the timeline below shows how it unfolded, with sources. Get the briefing to follow the top situations still developing: three emails a week, sourced and in order.
Unsubscribe anytime.
[SITUATION] · [QUIET] · [TECHNOLOGY]
118 clusters · 482 sources · 129 days · First seen · Last updated
Global AI infrastructure environmental and energy impacts
Overview
The rapid expansion of artificial intelligence is placing unprecedented strain on global electricity grids and water resources. To manage surging electricity demands, industry leaders like Nvidia, Google, and Microsoft are spearheading a shift toward 800-volt direct-current (800V DC) power architectures. On the energy supply side, the industry is increasingly looking toward nuclear power, though the reactivation of nuclear plants is expected to contribute less than 5% of the total increase in U.S. capacity.
In the United States, the demand for stable, continuous power has triggered a massive surge in gas-fired capacity. Moody’s Ratings estimates that approximately $110 billion in investment will be required to build 45 GW of new power generation capacity by 2030. Recent projections from Cleanview suggest U.S. operational data center capacity, currently at 61,560 megawatts, could increase sevenfold, supported by a pipeline of 2,142 projects aiming to add 380,312 MW. Texas is expected to lead this expansion, potentially capturing 75 to 100 GW of capacity.
To bypass utility connection delays, developers are increasingly turning to “behind-the-meter” projects, such as xAI’s mobile gas turbines in Tennessee and Fleet Data Centers’ temporary plants in Nevada. In Ohio, 12 proposed gas plant projects totaling 5,300 MW have sparked environmental concerns.
Addressing the carbon footprint of the hardware required for this expansion, Google has partnered with Stegra to support a new steel mill in Boden, Sweden. Located south of the Arctic Circle, the facility aims to use green hydrogen produced via water electrolysis powered by renewable wind and water energy. This process is estimated to reduce carbon emissions by up to 95% compared to conventional coal-fired methods. Under the partnership, Google will purchase environmental attribute certificates to claim these verified emissions reductions toward its climate goals.
Entities
Amazon · Texas · Google · Pacifico Energy · Pecos County
Claims
What the coverage asserts, and how many sources carry each claim.
- [● 7 SOURCES] The power plant is intended to operate independently of the public electrical grid.
- [○ 1 SOURCE] The planned data center campus is designed to transition to grid-connected service and will not raise electricity costs for Texas families.
- [● 12 SOURCES] Amazon is building a natural gas power plant in Pecos County, Texas, to power its AI data centers.
- [● 9 SOURCES] The facility is authorized to emit up to 33 million tons of CO2 annually.
- [● 7 SOURCES] The GW Ranch power plant is planned to feature 35 gas turbines.
- [● 7 SOURCES] The power plant is expected to have a capacity of up to 7.65 gigawatts.
- [● 6 SOURCES] Amazon's corporate emissions increased by 16 percent last year.
- [● 2 SOURCES] Construction of the plant is scheduled to begin in the first quarter of 2026.
- [● 2 SOURCES] Amazon projects capital expenditures of approximately 200 billion to 220 billion dollars for 2026.
- [● 2 SOURCES] Amazon's total backlog of orders reached 496 billion dollars.
- [○ 1 SOURCE] High energy demand and volatility in AI data centers can damage batteries, generators, and cooling systems.
- [○ 1 SOURCE] The GW Ranch project could reach over 5 GW of power generation by 2031 depending on customer needs.
- [○ 1 SOURCE] Amazon's AWS division reported a 37% year-over-year revenue increase in the second quarter.
- [○ 1 SOURCE] Amazon plans to use water not suitable for drinking or irrigation to avoid stressing the local watershed.
- [○ 1 SOURCE] The GW Ranch project will include 1.8 GW of battery storage and up to 750 MW of solar generation.
Timeline
-
[TECHNOLOGY] 15 sourcesGoogle backs Arctic hydrogen-based steel mill in Sweden
Google is partnering with Sweden-based Stegra to support a green hydrogen steel mill in Boden, aiming to offset the rising carbon footprint of AI-driven data center expansion.
-
[BUSINESS] 2 sourcesData center growth drives natural gas plant expansion in Nevada and Ohio
Rapid data center growth is driving the approval and proposal of new natural gas power plants in Nevada and Ohio to meet surging electricity demands.
-
[BUSINESS] 2 sourcesU.S. data center power capacity projected to increase sevenfold
U.S. data center power capacity is projected to increase sevenfold, driving a surge in electricity demand and a rise in on-site power generation to bypass outdated grid interconnection regulations.
-
[TECHNOLOGY] 2 sourcesUS data centers face rising gas consumption and fire risks
U.S. data centers are projected to consume more natural gas than most nations by 2035, while the rise of on-site power and battery storage increases fire and explosion risks tenfold.
-
[BUSINESS] 3 sourcesLNG demand rises due to AI growth and QatarEnergy supply shifts
AI-driven data center growth is set to boost Southeast Asian LNG demand, while QatarEnergy seeks long-term U.S. supply contracts to offset production losses.
-
[TECHNOLOGY] 2 sourcesNatural gas capacity surges to power U.S. AI data centers
The AI boom is driving massive growth in U.S. electricity demand, leading to a surge in natural gas power projects, particularly in Texas, to support expanding data center infrastructure.
-
[TECHNOLOGY] 5 sourcesData centers drive new natural gas demand in the US
AI-driven data center growth is driving new natural gas investments in the US, with companies like Amazon, Microsoft, and Nvidia seeking stable power to support massive energy demands.
-
[BUSINESS] 2 sourcesAmazon plans 7.65-GW gas power plant for Texas AI data center
Amazon plans to build a 7.65-GW gas power plant in Texas to provide dedicated energy for a new AI data center campus.
-
[TECHNOLOGY] 5 sourcesAmazon plans massive Texas AI data center and gas-fired power site
Amazon is developing the GW Ranch AI data center in Texas, featuring a massive gas-fired power plant. The project faces scrutiny over potential annual CO2 emissions of 33 million tons.
-
[TECHNOLOGY] 12 sourcesAI data center expansion drives shifts in power and infrastructure
AI growth is forcing a massive overhaul of data center infrastructure, driving shifts toward 800V DC power, nuclear energy, offshore cooling, and even orbital computing to meet unprecedented electricity demands
-
[BUSINESS] 3 sourcesAmazon to build massive Texas gas plant for AI data centers
Amazon is building a massive data center campus in Pecos County, Texas, powered by a dedicated 7.65-gigawatt natural gas plant that could become the largest single source of greenhouse gas emissions in the US.
-
[BUSINESS] 14 sourcesAmazon to build 7.65 GW gas plant in Texas for AI data centers
Amazon is building a 7.65 GW natural gas power plant in Pecos County, Texas, to power AI data centers, a project that could become a major US source of CO2 emissions.
-
[BUSINESS] 30 sourcesAmazon finances massive Texas gas plant for data centers
Amazon is financing a massive private gas power plant in Texas to power its AI data centers, a project that could become the largest single source of greenhouse gas emissions in the U.S.
-
[BUSINESS] 2 sourcesUS Data Centers Spur Rapid Build‑out of Natural‑Gas Power Plants
US data‑center expansion is driving a surge in natural‑gas power plants, with projects in Ohio, West Virginia and investments by Elon Musk, sparking political and environmental debate.
-
[TECHNOLOGY] 3 sourcesTech giants pour billions into nuclear reactors to power AI data centers
Microsoft, Amazon and Google are investing billions in small modular nuclear reactors for AI data centers, while Crusoe and Aalo Atomics aim to launch the first nuclear‑powered AI factory in Idaho by 2027.
-
[BUSINESS] 4 sourcesAI hyperscalers shift to nuclear and gas power for data center electricity
AI giants Microsoft, Google, Amazon and Meta have signed nuclear power deals totalling about 10 GW, while Meta quit RE100 after committing to gas plants for its AI data centres.
-
[TECHNOLOGY] 5 sourcesData Center Industry Faces Power and Workforce Shortages
AI data centers demand gigawatt‑scale power, causing multi‑year grid delays, while a shortage of 340,000 workers—especially veterans—hinders industry growth.
-
[TECHNOLOGY] 4 sourcesGoogle backs $4.5 bn Arkansas solar‑battery project for AI data centers
Google will fund a $4.5 bn Arkansas solar‑battery project delivering 2.5 GW solar and 2.9 GWh storage to power its AI data centers and 315,000 homes.
-
[BUSINESS] 3 sourcesVirginia Water Authority Clarifies Drought Measures and Google Data Center Water Use
Virginia officials urge heat safety and water conservation amid drought, while the Western Virginia Water Authority caps Google data‑center water use at 2 million gallons daily, debunking higher‑volume rumors.
-
[BUSINESS] 2 sourcesAI Fuels Surge in Energy and Rare Earth Demand, Prompting Major Investments
AI is boosting energy and rare‑earth demand, leading to large investments in power firms like Bloom Energy and increasing focus on rare‑earths for data‑center cooling.
-
[TECHNOLOGY] 4 sourcesOpenAI Data Center Uses Dry‑Cleaner Permit for Texas Power Plant
OpenAI’s Stargate AI data center in Abilene, Texas, built a natural‑gas power plant using a minor “dry‑cleaner” air permit, prompting local opposition over dust and emissions while officials cite economic gains
-
[TECHNOLOGY] 2 sourcesSpain moves ahead with major AI‑hydrogen hub in Huelva and sustainable hydroelectric pilot in Granada
Spain approves a €10 bn AI‑hydrogen hub in Huelva and launches HydroPulse to make Granada hydro plants more sustainable using AI and digital tools.
-
[TECHNOLOGY] 2 sourcesAI-driven predictive maintenance expands in European electricity grids
AI predictive‑maintenance tools are spreading in European electricity grids, but investment lags behind renewable growth; experts urge standards as Finland and Ireland pilot the technology.
-
[BUSINESS] 3 sourcesMitsubishi acquires $7.5 bn of U.S. natural gas assets from Aethon
Mitsubishi bought Aethon’s U.S. natural‑gas assets for $7.5 bn, becoming a top producer in the Haynesville Shale and targeting LNG exports to Japan and AI data‑center power demand.
-
[TECHNOLOGY] 15 sourcesAI data center boom strains power grids worldwide
AI‑focused data centres are pushing electricity demand to record levels, straining U.S. and Asian power grids, prompting project cancellations, and prompting firms like SpaceX and Hammond to seek new power‑sust
-
[BUSINESS] 8 sourcesGoogle secures full output of Arkansas' Steel River solar project to offset data‑center emissions
Google will buy all power from the first phases of Arkansas' Steel River solar project—1.6 GW solar, 1.9 GWh storage—to offset its AI data‑center emissions, part of a $3.5 bn effort targeting 24/7 carbon‑free C
-
[TECHNOLOGY] 3 sourcesMicrosoft datacenter consumes 1% of Dutch electricity, matching CO2 output of US cloud giants
Microsoft’s Middenmeer data centre used 1 % of Dutch electricity in 2023, while the combined CO2 emissions of Microsoft, Google and Amazon equal the Netherlands’ total output.
-
[TECHNOLOGY] 6 sourcesItaly passes unified data‑center law as Microsoft reports higher AI emissions
Italy adopts a unified data‑center licensing law while Microsoft reports a 25% rise in AI data‑center emissions, straining its carbon‑negative goal.
-
[BUSINESS] 5 sourcesU.S. gas‑plant boom spurred by AI data‑center power demand
AI data‑center demand is fueling a U.S. natural‑gas plant construction surge as utilities turn to gas while New York and other states push renewable‑use mandates for large data centers.
-
[TECHNOLOGY] 14 sourcesMicrosoft, Google and Amazon data‑centre build‑out spikes carbon emissions
Microsoft, Google and Amazon’s AI‑driven data‑centre boom pushed their combined emissions up 18‑25 % to 119 Mt CO₂e, challenging their net‑zero targets.
-
[TECHNOLOGY] 2 sourcesMitsubishi Estate launches $10 bn AI data‑center push in Japan
Mitsubishi Estate plans a 1.5 trillion‑yen, 2,500 MW data‑centre build in Japan as AI demand surges, raising power and environmental concerns.
-
[INTERNATIONAL] 2 sourcesUS Drives Nearly Half of Global CO2 Emissions Rise in 2025 as Fossil Fuels Remain Dominant
2025 data shows fossil fuels supplied 86% of global energy; the US caused 47% of the world’s emissions rise, driven by a 13% jump in coal‑power use and AI data‑center demand.
-
[TECHNOLOGY] 2 sourcesUS AI Data Centers Strain Water Supplies and Create Noise Pollution
AI data centers in the U.S. are using nearly a trillion liters of water annually and generating noisy hums that disturb residents, sparking environmental and health concerns.
-
[TECHNOLOGY] 2 sourcesAI adoption fuels security threats and soaring energy use
AI’s rapid adoption triggers fresh security threats—prompt injection, data poisoning, jailbreaks—and drives a 37% rise in electricity, 18% more CO₂ and 34% higher water use at Google, challenging both safety &
-
[BUSINESS] 2 sourcesEMEA data centre market faces power supply constraints
EMEA data centres see strong cloud/AI demand, but power and grid limits are curbing growth; operators explore self‑generation and microgrids as 7 GW of new capacity is built across major cities.
-
[TECHNOLOGY] 2 sourcesTencent releases Hy3 AI model with free access and wide product integration
Tencent launched its Hy3 AI model, offering a free, open‑source 295 B‑parameter system that outperforms rivals in benchmarks and is already deployed across many Tencent products.
-
[TECHNOLOGY] 22 sourcesAI Data Centers Spur Record U.S. Power Use and Global Grid Challenges
AI data centers are driving U.S. power demand to record highs, prompting grid warnings, emergency diesel use, community opposition, a Meta wastewater spill, and spurring Brazil’s hub ambitions amid regulatory &
-
[TECHNOLOGY] 3 sourcesMeituan open‑sources 1.6 trillion‑parameter LongCat‑2.0 model, Moore Thread adapts it for MTT S5000 GPU
Meituan open‑sources its 1.6 trillion‑parameter LongCat‑2.0 MoE model with 1 million‑token context; Moore Thread quickly adapts it for inference on its MTT S5000 GPU.
-
[TECHNOLOGY] 2 sourcesFerveret's Water-Free Adaptive Phase Cooling System for AI Data Centres
Ferveret, a MIT‑spun startup, offers a water‑free Adaptive Phase Cooling system that immerses servers in a specialised liquid, improving heat transfer and cutting energy use by about 15% for AI data centres.
-
[TECHNOLOGY] 7 sourcesNvidia unveils zero‑water cooling system for AI data centers
Nvidia’s new closed‑loop liquid‑cooling system for AI data centres eliminates direct water use and cuts cooling energy, saving costs while its net environmental benefit hinges on electricity sources.
-
[BUSINESS] 2 sourcesUS Fossil‑Fuel Power Investment Overtakes China Amid AI Data‑Center Boom
US fossil‑fuel electricity spending will top China’s in 2026, driven by AI data‑center demand and a policy shift toward gas and coal, while China expands renewables amid UN climate warnings.
-
[TECHNOLOGY] 13 sourcesKAIST study warns AI agents may demand half of U.S. electricity supply
KAIST finds AI agents use 136× more energy per query than chatbots, could need 198 GW (half US power) if scaled to billions of daily requests, highlighting urgent infrastructure challenges.
-
[TECHNOLOGY] 3 sourcesSpain's AI Data Centers Spur Energy Use, Local Opposition
Spain’s AI data‑centre boom raises electricity, water and housing concerns, mirroring protests in the US and Japan over similar facilities.
-
[TECHNOLOGY] 2 sourcesUnited Nations warns AI development outpaces regulation
The UN released its first global scientific AI assessment, saying AI’s rapid growth outpaces regulation and urging coordinated rules, while noting US and China dominate compute power.
-
[POLITICS] 2 sourcesSpain launches mandatory eco‑efficiency rules for tech companies
Spain's government unveiled the 'Vuelta a lo básico' plan on 2 July 2026, enforcing strict energy standards on tech firms, aiming for a 30 % cut in data‑centre power by 2028, with audits, sanctions and a €500 m
-
[TECHNOLOGY] 2 sourcesMeituan launches LongCat‑2.0, a trillion‑parameter AI model trained on domestic Chinese chips
Meituan released LongCat‑2.0, a trillion‑parameter AI model trained solely on Chinese chips, open‑sourced under MIT licence and benchmarked against top global models.
-
[TECHNOLOGY] 3 sourcesMilan taps data centre waste heat: A2A‑Equinix pact adds 225 GWh/yr
A2A and Equinix signed a pact to capture waste heat from a Milan data centre, adding 225 GWh/yr to the city’s district heating, enough for 21 000 homes and cutting 345 000 t CO₂.
-
[TECHNOLOGY] 2 sourcesArtificial Intelligence strains power grids and outpaces regulation, report says
AI data centres may use 3‑4% of global electricity, creating unpredictable grid loads, while fast‑moving AI advances outpace national regulation, raising safety and governance concerns.
-
[TECHNOLOGY] 3 sourcesGoogle AI expansion drives 37% jump in electricity use in 2025
Google’s 2025 electricity use jumped 37% due to AI data‑center growth, boosting water use and emissions, while the firm bought a record 12 GW of renewable power to offset impact.
-
[TECHNOLOGY] 14 sourcesGoogle and Amazon emissions spike as AI data‑centers expand
Google and Amazon report 82% and 58% emission rises since 2019, driven by AI data‑center expansion, jeopardising their net‑zero targets.
-
[TECHNOLOGY] 2 sourcesAI Industry Confronts Energy Grid Strain and Blockchain Integration Hurdles
AI data centers face grid strain, prompting battery storage and EV‑grid integration, while the blockchain‑AI sector battles demand, cost, and regulatory hurdles to adopt decentralized compute and storage.
-
[TECHNOLOGY] 2 sourcesSpain positions itself as AI data‑center hub amid rising energy and water concerns
Spain aims to become the AI data‑center hub of Southern Europe, but rapid growth raises energy, water and community‑impact concerns worldwide.
-
[TECHNOLOGY] 19 sourcesAI Data Centers Strain Power Grids as Compute Demand Surges
AI data centres are expanding rapidly, driving massive electricity demand that outpaces grid capacity, raising costs, spiking carbon emissions and prompting industry and policy concerns.
-
[TECHNOLOGY] 2 sourcesDenmark confronts AI‑driven data centre impacts and surge in municipal AI‑written complaints
Researchers warn Denmark’s expanding AI‑driven data centres will strain power, water and local communities, while a surge in AI‑generated citizen complaints is overloading municipal caseworkers.
-
[TECHNOLOGY] 3 sourcesAllianz Trade study warns data centers will become major electricity and carbon emitters
Allianz Trade warns data centers will double electricity demand, emit 286 Mt CO₂ in 2025—57 % above prior estimates—and could cost $154 bn annually by 2030.
-
[TECHNOLOGY] 2 sourcesMeituan launches LongCat-2.0, first trillion‑parameter AI model trained entirely on Chinese chips
Meituan unveiled LongCat‑2.0, a 1.6‑trillion‑parameter AI model trained entirely on Chinese chips, open‑sourced with low‑cost API access, highlighting China’s AI self‑sufficiency push.
-
[BUSINESS] 5 sourcesGerman towns convert former power plant and paper mill to battery storage and data centre
German towns repurpose old industrial sites: Hausham’s former gas plant will host a 44 MW battery storage unit, while Nidda’s ex‑paper mill is slated for a data centre that may use its water spring.
-
[TECHNOLOGY] 15 sourcesMeituan unveils 1.6‑trillion‑parameter LongCat‑2.0 model trained on Chinese chips
Meituan has open‑sourced LongCat‑2.0, a 1.6‑trillion‑parameter LLM with a 1 million‑token context, fully trained on a 50 k‑chip domestic ASIC cluster, aiming for self‑sufficient AI and claiming performance on a
-
[TECHNOLOGY] 3 sourcesChina to Synchronise Power Grid Expansion with AI Data‑Center Growth
China’s 2026‑2030 plan links renewable power expansion with AI data‑centre growth, targeting 4 TW capacity and over 90 % AI adoption by 2030.
-
[INTERNATIONAL] 3 sourcesUN urges global data‑center firms to disclose carbon impact and shift to renewable power by 2030
The UN has demanded that tech firms disclose data‑center emissions and transition all facilities to renewable energy by 2030, citing high electricity and water use.
-
[TECHNOLOGY] 2 sourcesNvidia unveils high‑temp liquid cooling to slash AI data‑center water use
Nvidia unveiled a high‑temperature closed‑loop liquid‑cooling system for AI data centres to cut huge water consumption, though rollout may take years.
-
[TECHNOLOGY] 2 sourcesJeff Bezos Water Quote on AI Proven Fabricated
A fabricated quote claiming Jeff Bezos prioritized water for AI over humans at VivaTech was debunked; it originated from a satirical account and spread widely on social media.
-
[BUSINESS] 2 sourcesIndia's power sector grapples with AI‑driven data‑center demand forecasting
AI‑driven data‑centres create “phantom” loads that complicate India's electricity forecasts, while firms turn to rooftop solar, storage and microgrids for reliable, low‑carbon power.
-
[INTERNATIONAL] 2 sourcesUN Announces Seven-Point Climate Roadmap at London Climate Action Week
UN Secretary‑General Guterres unveiled a seven‑point climate roadmap in London, urging major methane cuts, steep emissions reductions and new AI‑related environmental rules to achieve net‑zero by 2050.
-
[TECHNOLOGY] 3 sourcesChina's AI data centers confront green‑power shortfall
China aims for 80% renewable power in AI data centres by 2030, but grid technical limits, intermittent renewables and high‑peak AI loads make the target challenging; new AI‑focused power systems are being trial
-
[TECHNOLOGY] 2 sourcesVirginia drought forces data centers under same water restrictions
Virginia's severe drought has placed data centers under the same water-use limits as residents, with upcoming Google facility allowed up to 8 M gallons per day.
-
[TECHNOLOGY] 3 sourcesMicrosoft reports water‑positive data center operations for 2025
Microsoft says its 2025 data centers used less water than they replenished, advancing its water‑positive goal and introducing closed‑loop cooling that saves millions of litres per year.
-
[POLITICS] 3 sourcesAI data centers face US legislative push and UN call for environmental disclosures
US Congress debates AI data‑center cost measures while the UN urges firms to disclose environmental impacts and switch to renewables by 2030.
-
[TECHNOLOGY] 3 sourcesNvidia's Rubin AI servers use liquid cooling to cut data‑center water use to near zero
Nvidia's Rubin AI servers employ a closed‑loop liquid‑cooling system that eliminates fans, runs at up to 45 °C and can reduce data‑center water consumption to near zero while boosting performance and rack‑space
-
[TECHNOLOGY] 2 sourcesUN chief urges AI firms to disclose environmental footprint
UN Secretary‑General Antonio Guterres urged AI firms to disclose their energy, water and land use, warning that AI data centres could out‑consume most countries by 2030 and calling for renewable‑energy power.
-
[TECHNOLOGY] 2 sourcesNVIDIA unveils near‑zero‑water cooling for AI data centers
NVIDIA’s new liquid‑cooling system for AI data centers seeks to cut water use to near zero, aligning with tighter EU regulations but still facing climate and retrofit challenges.
-
[INTERNATIONAL] 2 sourcesUN Secretary‑General Guterres urges AI firms to disclose data‑center climate impact
UN chief Antonio Guterres, speaking in London, demanded AI companies reveal the climate impact of their data centers and urged a swift end to new fossil‑fuel investments to meet the 1.5 °C target.
-
[TECHNOLOGY] 37 sourcesItaly’s AI Push Highlights Workplace, Health and Sovereignty Challenges
Italy’s AI landscape faces workplace trust issues, health‑system digital gaps, sovereignty concerns, cybersecurity risks and calls for stronger EU research investment.
-
[TECHNOLOGY] 18 sourcesUN chief Guterres urges AI firms to disclose environmental impact
UN Secretary‑General Guterres urged AI firms to disclose carbon, water and land impacts of data‑centres and switch to renewable power by 2030, warning the sector could use more energy than all but five nations.
-
[INTERNATIONAL] 6 sourcesUN Secretary‑General Guterres launches global methane cut drive at London Climate Action Week
Guterres urged rapid methane cuts, launched a global Call to Action on methane and announced an AI environmental transparency initiative at London Climate Action Week.
-
[INTERNATIONAL] 2 sourcesLondon launches climate action week as UK heatwave intensifies
London hosts the 8th Climate Action Week with over 75,000 participants, UN Secretary‑General Guterres' seven‑point energy roadmap, and a city data hackathon, amid a UK heatwave.
-
[TECHNOLOGY] 3 sourcesData Center Waste Heat and Battery Storage Tapped to Power AI‑Driven Facilities
Berlin’s Ringberlin campus uses data‑centre waste heat for heating, while US AI data centres may double electricity use by 2030, prompting calls for large‑scale battery storage to balance the grid.
-
[TECHNOLOGY] 8 sourcesNvidia unveils liquid‑cooling design to cut data‑center water use
Nvidia's Rubin liquid‑cooling system for AI data centers runs at 45 °C, claims up to 100 % on‑site water reduction and lower energy use, but indirect water use from power generation remains high.
-
[BUSINESS] 2 sourcesAmazon achieves water‑positive status in India, returning more water than it uses
Amazon says it is now water‑positive in India, returning more water than it consumes in 2025 through reduced use, recycling and replenishment projects.
-
[TECHNOLOGY] 2 sourcesJeff Bezos AI‑water quote proven fabricated
A fake quote claiming Jeff Bezos prioritized AI water use over humans circulated after his VivaTech 2026 talk; fact‑checks confirm he never made such a statement.
-
[TECHNOLOGY] 3 sourcesUN report says AI growth could consume water equal to needs of 1.3 billion people
A UN report warns AI expansion could demand water for 1.3 billion people and generate 2.5 million tonnes of e‑waste annually, while other studies debate potential emissions offsets.
-
[TECHNOLOGY] 3 sourcesJeff Bezos warns water scarcity could limit AI data‑center expansion
Jeff Bezos said AI’s expansion needs massive water for data‑center cooling, warning that local scarcity could hinder development and sparking debate over AI’s environmental impact.
-
[TECHNOLOGY] 2 sourcesNobel Laureate Abhijit Banerjee Calls for Water Tax on AI Data Centers
Nobel economist Abhijit Banerjee urges a water tax for AI data centres, citing high consumption and the need for resource‑pricing policies.
-
[TECHNOLOGY] 2 sourcesAmazon declares water‑positive status for Indian operations
Amazon says its Indian operations are now water‑positive, returning more water than they use across offices, warehouses and data centres, a year ahead of schedule, as it expands AI investments in the country.
-
[TECHNOLOGY] 11 sourcesAI demand to drive massive global electricity surge up to 8% by 2050
Studies forecast AI will lift global electricity demand to 8% by 2050, with data‑center use rising 26% by 2026; projects span low‑energy chips in Argentina, China’s massive AI compute capacity, and US grid bott
-
[BUSINESS] 2 sourcesGlobal Data Centre Metal Demand Driven by Grid Infrastructure, Wood Mackenzie Says
Wood Mackenzie finds data‑centre metal demand is 3‑4× asset size due to grid and on‑site power needs, with aluminium peaking at 0.6‑0.9 Mt and power capacity rising to 30‑33 GW by the early 2030s.
-
[TECHNOLOGY] 3 sourcesGartner warns AI data‑center power demand could double by 2030
Gartner warns AI data‑centre electricity use could more than double by 2030, reaching over 1,200 TWh, prompting calls for massive grid investment and frugal AI approaches.
-
[TECHNOLOGY] 3 sourcesSpain’s data centre capacity forecast jumps 55% as global power use set to triple by 2030
Morgan Stanley lifts Spain’s data‑centre capacity forecast to 4.3 GW by 2035, spotlighting Aragón’s growth, while a global report says data‑centre electricity use will triple by 2030, driven by AI and cloud.
-
[TECHNOLOGY] 3 sourcesAmazon secures 20-year geothermal power deal for Nevada data centers
Amazon secures 20-year geothermal and solar‑battery power deal for Nevada data centers as U.S. AI facilities face drought‑linked water and power challenges.
-
[TECHNOLOGY] 2 sourcesAI’s soaring energy demand and regulation push spark debate in Denmark
AI’s rapid expansion raises concerns over energy use—equivalent to Belgium’s power demand—and spurs calls for stronger regulation as Denmark drafts new AI governance rules.
-
[TECHNOLOGY] 3 sourcesAmazon reports its data centres use seven times less water than industry average
Amazon says its 2025 data‑centre water use was 0.12 L/kWh – over seven times more efficient than the industry norm – with total consumption down 2 % despite growth.
-
[TECHNOLOGY] 2 sourcesAmazon AI Data Centers Use 9 Billion Litres of Water, UN Warns AI Could Consume 600 Billion Gallons by 2030
Amazon used 9 billion litres of water in its data centres last year and targets water positivity by 2030, while a UN report warns AI data centres could need 600 billion gallons of water globally by 2030, equal‑
-
[TECHNOLOGY] 2 sourcesAmazon reports 9.5 bn L water use in data centers, efficiency claim questioned
Amazon disclosed its data centers used 9.5 bn L of water in 2025, reporting a WUE of 0.12 L/kWh and claiming 7‑fold efficiency versus the industry, though the metric isn’t directly comparable. The firm aims for
-
[BUSINESS] 4 sourcesAmazon reports 2.5 billion gallons water used by data centers in 2025
Amazon says its data centres used 2.5 bn gallons of water in 2025, a 2 % drop from the prior year, achieving a 0.12 L/kWh usage rate—about seven times better than the industry average—and is 75 % toward its 203
-
[TECHNOLOGY] 3 sourcesGartner forecasts 26% rise in data centre electricity use by 2026
Gartner forecasts a 26% jump in global data‑centre electricity use by 2026, driven by AI servers, with total consumption potentially topping 1,200 TWh by 2030, raising grid‑capacity concerns.
-
[TECHNOLOGY] 2 sourcesUnited Nations warns AI could double global electricity use by 2030
UN forecasts AI will use 3 % of global electricity by 2030, demanding massive water, land and e‑waste, while businesses face limited emissions data from providers like Anthropic, affecting corporate carbon‑foot
-
[TECHNOLOGY] 2 sourcesUS AI data centers fuel water strain in drought‑hit regions
Two‑thirds of planned US AI data centers are slated for drought‑stricken areas, potentially raising water use to 73 billion gallons annually by 2028 and threatening local aquifers.
-
[TECHNOLOGY] 2 sourcesUN Report Warns Artificial Intelligence Could Consume Water for 1.3 Billion People by 2030
A UN report warns AI could use water for 1.3 bn people and 945 TWh of electricity by 2030, making it the 11th‑largest power consumer globally, with vast variations in task energy use.
-
[TECHNOLOGY] 2 sourcesUN Report: AI Data Centres Could Use Water Equal to Needs of 1.3 Billion People by 2030
UN report warns AI data centres may need water for 1.3 billion people and 945 TWh electricity by 2030, urging broader environmental assessment.
-
[TECHNOLOGY] 2 sourcesUN report warns AI data centers could use water equivalent to 1.3 billion people by 2030
UN study projects AI data centres will need water for 1.3 bn people and 945 TWh electricity by 2030, requiring 14,500 km² of land and generating massive carbon and waste footprints.
Sources
199it.com · 24hod.sk · 24matins.fr · 31mag.nl · 36kr.com · 4sysops.com · 919thebend.ca · abc17news.com · abmedia.io · actualidadecommerce.com · africanewswire.net · agendadigitale.eu · agendarweb.com.ar · aijourn.com · almanzora.ideal.es · almouwatin.com · altonivel.com.mx · americafirstreport.com · amp.ilgiornale.it · analyticsinsight.net · anatolh.com · anda.jor.br · android-mt.ouest-france.fr · anglophone-direct.com · aop.com.br · apcmag.com · apre.it · archynetys.com · arroyodiario.com.ar · asiae.co.kr · autoblog.com.ar · aviaanaccounting.com · avvisatore.it · awreferencement.com · babyboomster.com · baoquangtri.vn · barlamantoday.com · benmilne.com · bezva-rady.cz · bhaskarlive.in · bigdatamagazine.es · binaryboxstudios.com · bitmat.it · biz.heraldcorp.com · blockmedia.co.kr · blog.apc.com · blog.teamascend.com · blogs.nvidia.cn
This summary has been updated 14 times: see revision history