The “cloud” is often thought of as a weightless, invisible sort of thing floating around somewhere out of sight. The reality, however, is a lot more grounded and resource-heavy. Data centres are seen as the physical engine rooms of the Internet and as our digital lives become more demanding, their appetite for power and water is increasing too.
Every time you ask an AI chatbot a question, stream a movie or even just send an email, servers in these centres fire up to handle your request. And you can bet they get hot and thirsty while doing it.
Globally, data centres already account for approximately 1.5% of the world’s electricity usage. In 2024, that was roughly 415 terawatt-hours a year and it’s rather concerning that that figure is expected – and is on track – to double by 2030.
Then there’s water. A single large facility can get through around five million gallons a day which is about the same as a town of 50,000 people. Most of it is used to stop the equipment from overheating. If you can imagine how many people use the Internet daily, now even more so with AI, the prevention of overheating is necessary.
But where does that leave us? Is it just the unavoidable price of progress or is there a more sustainable way to keep the lights on? Industry experts have shared their thoughts on the real environmental cost of data centres.
Our Experts
- Patrick Keaney: Global Engineering Executive at SLR
- Louis Leung: Founder of inFlow Inventory
- Oliver Moheda: Music Manager & Founder of Total Artist Management
- Kumar Brahnmath: Chief Product & Technology Officer at Measurabl
- Polly Milne: Chief Operating Officer at FuturePlus
- Mark Rushworth: CEO & Co-founder of Finchetto
- Daniel Hunt: UK Country Head at Agilité
- Dawn Barclay‑Ross: AI‑Driven Events Founder & Event Director, FUND Expo 2026
- Anthony Abate: Vice President & Chief Technology Officer at AtmosAir Solutions
- Mark Skelton: Chief Technology Officer at Node4
- Damir Špoljarič: Managing Partner at Gi21 Capital
Patrick Keaney, Global Engineering Executive at SLR
“To be net positive on water, the tech sector has to move past just being more efficient and start taking real accountability for its impact. That starts with measuring water the right way – at the basin level, not just total consumption – and shifting toward non-potable and recycled sources wherever possible.
It also means putting water back into the system through replenishment, not just reducing use. But the real unlock is how AI gets applied. If it’s focused on solving meaningful water problems – reducing losses in utilities, improving industrial reuse, optimising agriculture, the impact can far outweigh the sector’s own footprint.
A few companies are starting to head in this direction with water positive commitments, but most are still early. The ones that will get there are the ones that treat water as a real operating constraint and aim their technology at problems that actually move the needle.”
Louis Leung, Founder of inFlow Inventory
“I can say that software efficiency remains a neglected factor in the analysis of the environmental costs incurred by data centres. Sustainable conversations often revolve around renewable energy, cooling infrastructure, or server hardware. Despite the interconnected systems of modern cloud platforms, it is the SOFTWARE DESIGN that determines how much of those resources each resource should consume.
It is crucial, however, to treat sustainability as a performance metric. It is common for development teams to track uptime, latency, and fault tolerance, HOWEVER they rarely measure the amount of compute needed to achieve these goals. Adding computational efficiency to architecture reviews and load testing can absolutely reduce resource consumption without sacrificing performance.
Data storage optimisation, minimisation of duplication/ redundancy, and cost-efficient scaling drastically reduce demands on infrastructure without compromising core functionality. While data centres can and must become greener, we also need to complement it with software that make better use of those resources.”
Oliver Moheda, Music Manager & Founder of Total Artist Management
“The environmental cost of data centres has become one of the most significant challenges facing the digital economy. As AI adoption, cloud computing and streaming services continue to grow, demand for data processing is increasing at an unprecedented rate, placing pressure on energy grids and water resources worldwide.
However, the industry is not standing still. Many operators are investing heavily in renewable energy agreements, more efficient cooling technologies and next-generation infrastructure designed to reduce power consumption. We are also seeing growing interest in locating facilities in regions where renewable energy is abundant and temperatures naturally support more efficient cooling.
From the perspective of the music and entertainment sector, digital services are essential to how artists connect with audiences, distribute content and build careers. The challenge is ensuring that technological growth remains sustainable. Businesses, technology providers and policymakers must work together to balance innovation with environmental responsibility.
The future of data centres will not be defined solely by processing power, but by how efficiently that power is delivered. Organisations that prioritise sustainability today are likely to gain both environmental and commercial advantages in the years ahead.”
Kumar Brahnmath, Chief Product & Technology Officer at Measurabl
“Data centres are now among the most energy-intensive assets on the planet, and demand is accelerating fast. The proliferation of AI workloads, real-time analytics and cloud-native infrastructure has pushed data centre energy consumption to the top of the corporate sustainability agenda — and rightly so. For organisations with net zero commitments, the carbon embedded in their digital infrastructure is no longer an afterthought; it’s a material exposure.
The industry is responding, but unevenly. We’re seeing meaningful progress on renewable energy procurement and cooling efficiency — liquid cooling and waste heat recovery are moving from pilot to mainstream. But the harder challenge is transparency. Many organisations still lack granular, reliable data on the actual energy consumption and carbon intensity of the data centre assets underpinning their operations, particularly where infrastructure is outsourced or co-located.
That’s where measurement becomes critical. You can’t decarbonise what you can’t see. Platforms that bring data centre energy use into a unified ESG data framework — alongside buildings, transport and supply chain — give organisations the visibility to act. The environmental cost of digital infrastructure is real and growing. The question is whether businesses are equipped to account for it accurately, and manage it strategically.”
Polly Milne, Chief Operating Officer at FuturePlus
“Data centres account for roughly 1.5% of global electricity consumption, and that share is rising as AI workloads grow. For many businesses, the emissions tied to cloud services, SaaS, storage, and AI sit in Scope 3 as purchased digital services, often without much visibility. Most SMEs are not tracking the carbon footprint of their digital operations in any meaningful way.
The industry is responding, with hyperscalers including Microsoft and Google making major commitments on renewable electricity sourcing and water efficiency. The EU now requires large data centres to report energy-performance data. But commitments and operational reality on the ground are not the same thing. Matching electricity use with 24/7 carbon-free energy remains difficult for most operators, and rapid AI growth is still outpacing many efficiency gains.
What businesses can do now is ask the right questions of their tech suppliers: What is your PUE (Power Usage Effectiveness)? How do you source electricity, and is it backed by credible renewable procurement? Do you publish sustainability data or a verified sustainability report? Are you connecting environmental sustainability with social impact on the ground? Procurement is one of the most underused levers in corporate sustainability, and digital procurement is no exception.”
Mark Rushworth, CEO & Co-founder of Finchetto
“The data centre of the future can’t just be bigger; it has to be smarter. That means rethinking power distribution, thermal design, and especially network efficiency. Every week, we read forecasts all pointing to the exponential growth of AI and the growing strain on environmental resources.
What we need is short to medium-term solutions to curb the curve, and that is through investing in novel types of hardware that deliver the performance needed for AI but consume far less power. Photonics offers a rare opportunity to leapfrog, not just iterate, on the limitations of electrical switching. It’s a physics-led path to both performance and efficiency.”
Daniel Hunt, UK Country Head at Agilité
“The environmental cost of data centres is real, but it cannot be understood through energy use alone. These are highly technical buildings, shaped by power demand, cooling requirements, construction complexity, supply chain pressure and the need to adapt as technology changes.
As AI increases rack density, the industry is having to think differently about the physical design of these environments. More power-intensive equipment changes the cooling strategy, the plant requirements and the way space is planned around future flexibility. Liquid and hybrid cooling are becoming part of that conversation, but the wider challenge is to make sure sustainability is considered early, rather than treated as something to solve later.
One of the most interesting opportunities is heat reuse. In the right location, surplus heat from data centres can support local infrastructure, helping these buildings move from being seen only as heavy energy users to more useful parts of the built environment. The industry is also becoming more conscious of waste and adaptability.
If data centres are designed to respond to change, rather than being repeatedly stripped out and reworked, they can become more resilient, more efficient and more responsible over time.”
Dawn Barclay‑Ross, AI‑Driven Events Founder & Event Director, FUND Expo 2026
“Most coverage of data centres focuses on a very one‑dimensional story: “they’re energy‑hungry and bad for the planet.” The reality is more complicated – and more hopeful.
Yes, large data centres consume significant electricity and water, and that has to be taken seriously. But we also need to recognise that they now underpin almost everything we value in a modern economy: AI, cloud services, payments, logistics, telemedicine, education and more. Switching that infrastructure off is not an option; making it smarter is.
The real opportunity lies in treating data centres as part of a wider system rather than as isolated “energy drains”. When you start coupling them with waste‑to‑energy, district heating and local grid balancing, you can turn heat and by‑products that would otherwise be wasted into something genuinely useful – warming homes, powering nearby industry, stabilising local networks.
The goal shouldn’t be “no data centres”, but data centres that pay their way environmentally as well as economically.”
Anthony Abate, Vice President & Chief Technology Officer at AtmosAir Solutions
“Data centres can have a significant environmental impact because they consume enormous amounts of electricity to power servers and the cooling systems needed to prevent overheating. Data centre growth is increasing demand for energy and water resources, raising concerns about greenhouse gas emissions, strain on local power grids, and sustainability.
One way to reduce energy and lower the environmental risks starts on the roof by upgrading new or existing heating and air conditioning (HVAC) technologies, such as adding DBD bi-polar ionisation technology to data centre HVAC systems.
BBD BPI devices help reduce the amount of outside air required, making data centres run on optimum conditions, lowering both energy usage and greenhouse gas emissions making them more sustainable. Keeping the air pristine inside data centre facilities is key to operating at full proficiency while minimising HVAC energy use which can increase environmental hazards that can possibly impact local communities.
Upgrading HVAC technology should be a key consideration for architects, builders and those companies that are building or upgrading data centres.”
Mark Skelton, Chief Technology Officer at Node4
“In the UK, data centres already consume around 2% of national power, while unchecked growth could push that to 10–15%, at a time when the grid is already strained. There’s also a broader question of how well prepared the UK actually is for such a rapid scale-up in data centre infrastructure.
We have witnessed a noticeable shift in priorities when it comes to ESG initiatives over the past few years. As AI-driven growth accelerates, ESG and sustainability appear to be slipping down the agenda. Some organisations are even scaling back ESG functions in the race to deploy new technologies. But that’s a mistake.
AI and sustainability are not and do not have to be opposing forces. If anything, they should evolve together. The rapid expansion of compute capacity, data centres and digital infrastructure brings real environmental implications. Without clear guardrails and coordinated oversight, we risk solving one productivity challenge while creating a long-term energy problem.
This is where government leadership matters. Investment in AI must sit alongside investment in sustainable energy, grid modernisation, and efficiency standards. Growth at any cost is not a strategy. Sustainable growth, underpinned by infrastructure and governance, is.”
Damir Špoljarič, Managing Partner at Gi21 Capital
“Data centres have become an easy villain, but most of the criticism is aimed at the wrong target. Put the water panic in perspective: California’s almond farms use close to a hundred times more water than all of America’s data centres, and golf courses around thirty times more. Nobody’s marching on the ninth hole.
That said, operators don’t get to hide behind a good chart. The impact is local and concentrated, and AI is scaling demand fast, so “better than almonds” is not a sustainability strategy.
The encouraging part is that the industry is already moving. Closed-loop and direct-to-chip liquid cooling are pushing water use toward zero. Waste heat that used to be vented to the sky is being piped into homes and greenhouses. And the smartest new capacity is being built where the power is already renewable and the climate does the cooling for free.
That’s the real answer to keeping up with AI without wrecking the planet: you don’t bolt AI onto twenty-year-old buildings designed for handling emails. You build new, dense, efficient capacity in the right place. The footprint problem isn’t the data centre, it’s the wrong data centre, in the wrong location, on the wrong power.”
